1
50
1
-
https://omeka.library.appstate.edu/files/original/011d930e1aaa90fc550d435a33553c19.pdf
aded58e7472ed181e302709aa926eba5
PDF Text
Text
�People, Politics and Economic Life
Exploring Appalachia with Quantitative Methods
Updated Second Edition
Thomas Plaut
With an overview of the Appalachian Region by Susan Emley Keefe
KENDALL/HUNT PUBLISHING COMPANY
in cooperation with
APPALACHIAN CONSORTIUM PRESS
Software created by MicroCase Corporation
31/2 inch IBM-compatible diskette enclosed
KENDALL/HUNT PUBLISHING COMPANY
4050 W es t ma r k Drive
Dubuque,
Iowa 52002
�The Appalachian Consortium was a non-profit educational organization
composed of institutions and agencies located in Southern Appalachia. From
1973 to 2,004.; its members published pioneering works in Appalachian studies
documenting the history and cultural heritage of the region. The Appalachian
Consortium Press was the first publisher devoted solely to the region and many of
the works it published remain seminal in the field to this day.
With funding from the Andrew W. Mellon Foundation and the National
Endowment for the Humanities through the Humanities Open Book Program.;
Appalachian State University has published new paperback and open access
digital editions of works from the Appalachian Consortium Press.
www.collections.library.appstate.edu/appconsortiumbooks
This work is licensed under a Creative Commons BY-NC-ND license. To view a
copy of the license^ visit http://creativecommons.org/licenses.
Original copyright © 1996 by the Appalachian Consortium Press.
ISBN (pbk.: alk. Paper): 978-1-4696-4134-8
ISBN (ebook): 978-1-4696-4135-5
Distributed by the University of North Carolina Press
www.uncpress.org
�CONTENTS
I.
Susan Emley Keefe: Appalachia and Its People
3
II.
Exploring a Region through Quantitative Data
Frequencies and Percentages
25
Cultural Diversity in Appalachia
Mapping, Rank Ordering and Correlation Coefficients
34
Demographics: Urbanization and Migration in Appalachia
Mapping and Regression
45
The Appalachian Economy I: Comparing Subregions
Analysis of Variance (ANOVA) in Group Scores
57
The Economy II: Industry and Opportunity
Correlations and Regressions
75
Voting Patterns and Economic Conditions
Tables, % , and Nonparametric Measures of Association
85
III.
IV.
V.
VI.
VII.
VIII.
Race and Region: Minorities in Appalachia
Multiple Regression
101
IX.
A Region with Many Definitions
114
X.
The Internet: Finding New Data in The Information Age
125
Glossary
133
Codebook for the APCOUNTY File Variables
139
Suggested Readings and Media Resources by Chapter
145
ill
�Preface to the Second Edition
This edition updates the data set and responds to the insights of users over the past three years.
Students, both undergraduate and graduate, report finding the book to be a good means of learning
or relearning and reviewing statistical methods and consequently I've attempted to refine the
commentary on data analysis. Faculty report that students can work through the material with
minimal guidance and thus are prepared to talk about results—and the questions raised by the results—
in class. New exercises have been added for summarizing research results, which make use of
software not available at the time of the first printing (such as Microsoft's PowerPoint or Corel's
Presentations).
There is a new chapter on the Appalachia as a region, which serves to balance and explicate the
quantitative material and exercises. The chapter was written by Susan Emley Keefe, chair of the
Department of Anthropology at Appalachian State University, who has numerous publications on
the region to her credit, including Appalachian Mental Health (University Press of Kentucky, 1988).
Definitions of Appalachia have been continually changing over the past century. In his chapter on
"Concept and Method" in Glen E. Lich's Regional Studies: The Interplay of Land and People
(College Station, Texas, Texas A&M University Press, 1992), Terry G. Jordan describes three
concepts of region: 1) the formal, having a homogeneity of traits; 2) the vernacular, which has a
"broadly perceived regional self-consciousness," and 3) thefunctional. The functional region "is an
area that has been organized to function politically, socially or economically" (page 20). Sue Keefe's
chapter provides the evidence for Appalachia as a formal region, which some might also argue
contains evidence for a vernacular one. Appalachia, as defined by the Appalachian Regional
Commission, is best described as a functional region.
There is a new chapter on the Internet to walk students through obtaining information at the national,
regional and county levels. The Internet chapter provides links to public and private agencies,
educational institutions and citizens' groups involved in the Appalachian region. A data file adapted
from this chapter is included on the disk accompanying the book. Users can sign onto the Internet,
then pull up either the WordPerfect or Microsoft Word "Internet launcher" file and click on the blue
hyperlinks to access web sites described in the text.
The rapid changes of the Information Age created some challenges for this edition. There was new
data to be incorporated since the first publication three years ago. Six new variables on population,
income and poverty have been added to the data set. The mapping software could not be changed and
consequently seven counties added by Appalachian Regional Commission to its definition of the
region in 1998 are missing: Macon and Hale counties in Alabama; Elbert, Hart and Yalobusha in
Georgia, and Montgomery and Rockbridge in Virginia. The chapter on the Internet was designed in
part to overcome this problem; users can access the Commission's web site for data on these and
other counties.
IV
�A note on software: Since this student version of the MicroCase software is DOS based, Windows
users can move out of the program into other applications by holding down the Alt key and pressing
the Tab key. The program can be entered again with a mouse click on the bar at the bottom of the
monitor screen. The APCOUNTY data file can be accessed by the complete MicroCase Analysis
System statistical package available in many colleges and universities. This package provides a full
range of statistical procedures and the ability to convert the file for use by other statistical programs
such as SPSS.
October 2001
Preface
Appalachia often has been portrayed as a distant mountain region inhabited by less-than-reputable
holdovers from the American frontier who somehow escaped the blessings of modernization. This
stereotype sadly misses the mark. Extending from New York to Mississippi, Appalachia is as varied
as any region in the United States, containing great differences in geography, climate and culture.
Like many places in America, it includes areas facing economic hardship as well as areas enjoying
growth in industry, population and per capita income. It has both wealth and poverty, isolated rural
hamlets and metropolitan areas. Each year, thousands of people hike its mountain trails, raft its
Whitewater streams and pack into its music and storytelling festivals.
As a region of contrasts, Appalachia can be viewed as a "slice of the American pie." Yet its
geography and resources have also made it unique. The coal industry has dominated the central part
of the region with widespread impacts. A long tradition of mountain family farming, especially in
Southern Appalachia, has emphasized the importance of self sufficiency, social equality and the
bonds of kinship and community. The region has been the target of a number of experiments in social
change and engineering, from turn-of-the-century missionaries to the poverty warriors, dollars and
politics turned loose in the Appalachian Regional Development Act of 1960s. Size is significant:
Appalachia includes 399 counties (12.7 percent of all U.S. counties) in 13 states (26 percent of all
the states). It is big enough to reflect the complexities of late 20th century America and yet small
enough to retain its own history and cultural identity. It provides a wealth of data for people
interested in regional history, economics, modernization and social change.
This workbook is designed to provide the means for exploring issues facing any American region,
as well as Appalachia. It can be used as a supporting text in courses on regional studies, geography,
history and the social sciences, or simply as a source for people thinking about their own
communities. Readers will work with data from the U.S. Bureau of the Census, Bureau of Economic
Analysis, Appalachian Regional Commission and other agencies. A computer disk containing the
necessary software is included so that, as the book raises questions, the reader can find answers on
any IBM-compatible computer with 640K of memory (RAM) and a VGA graphics card. The disk
contains a data file with 96 variables for the 399 Appalachian counties, along with a trimmed-down
version of the MicroCase statistics package.
May 1996
�Acknowledgments
A number of people have assisted in the development of this volume. Ron Eller and Gordon
McKinney, who direct Appalachian Centers at the University of Kentucky and Berea College,
respectively, reviewed the text from the perspective of their years of research and work in the
mountains. Marian Plaut brought her talents as an editor, therapist and spouse to the manuscript,
as did June Trevor, a colleague at Western North Carolina Community Health Research Services
at the Mountain Area Health Education Center in Asheville. Mars Hill College President Fred
Bentley offered funding for research and publication of the workbook. John Payne, the college's
Dean for Learning Resources, funded the development of the book's computer map and the
research that linked the county-by-county data to it. The students who beta-tested the exercises
were Jamie Willis, Robert Gouge, Joy Greene, Josh Sparks, Sherry Wagner, Condalisa Hankton,
Marne Crisp, Hope Lloyd, Alicia Payne and Kara Hensley. Phil Obermiller, Dick Couto, Ray
Rapp, Walter Stroud and Ken Sanchagrin provided counsel and support at various stages in the
development of this project. Judy Slater, Laurie Pederson and George Peery reviewed the
manuscript for the second edition.
William Fox reviewed the manuscript with humor and a sharp eye for statistics. His text, Social
Statistics using MicroCase, was a primary tool in my developing the manuscript; its glossary
served as a template for my own. The Appalachian Regional Commission's Salim Kablawi, Judith
Maher and Ann Anderson were helpful in accessing the latest available data for Appalachia's 399
counties. MicroCase Corporation's David Smetters arranged for the production of the
computerized map which is so central to the book and patiently gave advice and counsel in the use
of his company's statistical analysis package.
The Author
Thomas Plaut is a Professor of Sociology and Director of the Center for Assessment and Research
Alliances (CARA) at Mars Hill College, a small liberal arts college founded by farmers for their
children in the mountains of Western North Carolina near Asheville in the mid nineteenth century.
He is a consultant and trainer for hospitals, hospices and health systems and has authored articles
on research methods, regional issues and cultural differences between service providers and rural
clients and staff. He received his A.B. degree from Harvard, M.A. from American University and
Ph.D. from the Union Graduate School.
VI
�This page intentionally left blank
�For Marian
�Appalachia as defined by the
Appalachian Regional Commission in 1996:
399 counties in 13 states
�This page intentionally left blank
�I. Appalachia and Its People
Susan Emley Keefe
Appalachian State University
Defining the Region
The name "Appalachia" originates in the
geography of the region. The most distinctive regional
feature is the Appalachian Mountain chain, which was
named by early Spanish explorers who borrowed the
term from an Indian tribe, the Appalachee, in northern
Florida (Raitz and Ulack 1991), Appalachia consists of
four physiographic regions: the eastern piedmont, the
Blue Ridge Mountains, the Greater Appalachian
Valley, and the Allegheny-Cumberland Plateau. The
Blue Ridge range was created almost 500 million years
ago when land masses from east and west moved
together and the stress caused the earth to buckle into
parallel ridges. The Alleghenies are newer. There was
a shallow sea where the Greater Appalachian Valley is
now. The sea receded and the earth beneath buckled,
creating the upland plateau area. The north-south
Appalachian mountain chain, stretching from Alabama
to New York, makes east-west travel difficult,
heightening the importance of the Greater Appalachian
Valley as a transportation route through the region.
As a geographic and cultural area within a nation,
there are no fixed boundaries defining Appalachia.
Geographers tend to be inclusive, extending the
regional boundaries to the edges of the piedmont and
plateau provinces. They often vary, nevertheless, in
marking the exact regional boundaries, taking into
consideration differences in elevation, climate, soil
type, and cultural environment (Raitz and Ulack 1991).
On the other hand, sociocultural studies tend to put
greater emphasis on cultural features, emphasizing a
core area in the southern highlands. In his classic
This chapter has been adapted from
"Appalachian Americans: The Formation of
Reluctant Ethnics" in Gregory R.
Campbell, ed., Many Americas; Critical
Perspectives on Race. Racism and
Ethnicity. Dubuque, 10: Kendall/Hunt 1998.
study, The Southern Highlander and His Homeland
(1921), John C. Campbell included West Virginia and
portions of 8 other states; Alabama, Georgia, South
Carolina, North Carolina, Tennessee, Kentucky,
Virginia, and Maryland. He called the region bounded
by natural topography "a land of mountains, valleys,
and plateaus." He also noted its physical isolation,
culturally homogeneous heritage, and stable
population.
In a notable economic-based survey of the region in
1962, The Southern Appalachian Region: A Survey,
Thomas Ford reduced Campbell's region to the
heartland, including West Virginia and portions of 6
other states: Alabama, Georgia, North Carolina,
Tennessee, Kentucky, and Virginia. While Campbell
emphasized culture and geography in defining the
region, Ford focused on the lack of economic
development as a defining characteristic.
In 1965, the federal government established the
Appalachian Regional Commission (ARC) and defined
the Appalachian region by the criterion of economic
need. West Virginia and portions of 12 other states
(Mississippi, Alabama, Georgia, South Carolina, North
Carolina, Tennessee, Kentucky, Virginia, Maryland,
Ohio, Pennsylvania, and New York) are eligible for
monies from the multi-billion dollar economic
development program established by the Appalachian
Regional Development Act.
Political maneuvering by governors hoping to include
underdeveloped portions of their states added areas
never before considered part of the Appalachian region.
The primary considerations for inclusion were, finally,
geographic contiguity and economic underdevelopment
marked by lack of jobs and forced migration,
overspecialization of the economy, a high rate of
poverty, low rates of education, and poor health
facilities. The purpose of the ARC is to rectify these
conditions with highway construction, health services,
educational and sanitation facilities, and small business
expansion. As defined by the ARC, Appalachia has
26.5 million people, which is 11.7% of the US
population (US Bureau of the Census 1980).
�Recognizing the diversity encompassed within the
borders of its definition of Appalachia, the ARC
identifies three subregions: Northern, Central, and
Southern Appalachia (ARC 1965).
Northern
Appalachia, the largest of the three, includes designated
counties in New York, Pennsylvania, Ohio, Maryland,
and northern West Virginia. It is also the most
populous and the most urban, taking in seven US
Census-designated Standard Metropolitan Statistical
Areas (SMSAs) including Erie, PA, and Binghamton,
NY. Northern Appalachia is an old industrialized area
based on nineteenth century coal, steel, and railroad
development which declined in prominence following
the Depression and the growth of highway
transportation in other parts of the country. The older
factories became obsolete, cities began to deteriorate
physically, and many people migrated out.
Central Appalachia is composed of counties in
eastern Kentucky, northwestern Tennessee, southern
West Virginia, and southwestern Virginia. This
subregion encompasses the smallest area and has the
smallest population. It is also the most rural subregion,
and includes no US Census SMSAs. It has the poorest
population because it is dominated by a single-resource
economy: coal. There are rich deposits of bituminous
coal, and Central Appalachia holds the main US
deposit of anthracite coal. The region's economy
fluctuates with the coal economy resulting in boomand-bust cycles which, along with increasing
mechanization in the mines, have created huge waves
of out-migration, especially in the mid-twentieth
century.
Southern Appalachia takes in counties in eastern
Tennessee, southwestern Virginia, North Carolina,
South Carolina, Georgia, Alabama and Mississippi. It
is the most rapidly growing subregion due to an
expanding economy based on industry, recreation, and
tourism. Historically an agricultural and timber region,
industrialization came relatively late to Southern
Appalachia in the last half of the twentieth century.
Growing urbanization and its associated social
problems have affected the subregion, which contains
six US Census SMSAs, including Knoxville, TN, and
Asheville, NC. Some of the nation" s most visited parks
and forest areas are located in Southern Appalachia,
including the Blue Ridge Parkway and the Great
Smokey Mountain National Park.
While Appalachian scholars recognize the shared
economic conditions in these three regions and
similarities in way of life stemming from a lack of
resources, most general studies about Appalachian
people and culture focus on the more homogeneous
population residing below the Mason-Dixon line, that
is, what ARC designates as Central and Southern
Appalachia. It is this area that is emphasized in the
following.
Racial and Cultural Diversity
The Appalachian region was originally populated, of
course, by indigenous Native Americans. By the time
of European contact, tribal groups in the southeastern
United States were extensive in number and had
achieved a high level of social organization. It is
estimated that as many as 60,000 Indians lived in the
Appalachian mountains at the time of contact. As in
other areas of the New World, these native populations
were decimated soon after discovery due to warfare,
forced labor, and disease. The discriminatory policies
of the US government in the 18th and 19th centuries,
and the forced removal of many groups, such as the
Cherokee on the Trail of Tears to Oklahoma, resulted
in further population declines. Today, relatively few
Native Americans live in Appalachia. The Qualla
Boundary in North Carolina, principal reservation of
the Eastern Band of Cherokee Indians, has the largest
concentration with a contemporary population of
11,500. In some areas, pockets of mixed-blood peoples
of Indian-African-European heritage remain, such as
the Melungeons in eastern Tennessee. Their origin and
racial classification continue to be problematic (Beaver
and Wilson 1997).
African Americans have been present in the
mountains since the 1500s, brought first as slaves of
Spanish and French explorers (Cabbell 1985). The
mountains of Appalachia offered refuge for small
numbers of freedmen and escaped slaves before the
Civil War, and it was through mountain routes that the
Underground Railroad transported thousands of
runaway slaves to northern ports and Canada during the
mid-1800s. Slavery, of course, was present in the
mountains, but there is evidence that its character was
unique (Inscoe 1989). Due to the absence of a
plantation system of agriculture in the mountains,
slaves were less numerous than in other parts of the
south. Appalachian homesteads were relatively small,
subsistence-based family farms with little need for a
large labor force, and, in fact, only about 10% of the
households in Appalachia held slaves. Where slaves
were owned, they were often used in non-agricultural
labor, as hotel servants or blacksmiths, and they were
given somewhat more freedom in their movements than
slaves in the plantation South. They were often
�household slaves, sharing living accomodations with
the master's family. There is some evidence, perhaps
due to these unique conditions of slavery, that white
attitudes towards blacks were less prejudiced during
this period in Appalachia than elsewhere in the Deep
South (Cabbell 1985).
Following Emancipation, the demand for coal miners
far exceeded the supply in Kentucky, Tennessee,
Virginia, and West Virginia, and consequently there
was a major migration of blacks from Alabama and, to
a lesser extent, Mississippi. In 1860, only about 15,000
blacks lived in Central Appalachia. But by 1920 almost
90,000 blacks worked in the coal fields, and 69% of
them resided in West Virginia where the full range of
Jim Crow laws had never been enacted, in contrast to
neighboring Appalachian states (Lewis 1987). Blacks
chose to work in the Appalachian coal mines for sound
economic reasons. As Ronald Lewis states, "Blacks
not only were welcomed in the mountain coalfields,
they were given equal wages for equal work and as
good an opportunity in the occupational hierarchy as
they were likely to find anywhere in industrial
America" (1987:143). Certainly the life of a miner was
difficult, but it was equally difficult for blacks and
whites. Socially, however, blacks remained segregated
in almost all coal company towns, where they had their
own neighborhoods, recreational facilities, companysponsored baseball teams, and schools.
The mines also attracted eastern and southern
European immigrants, particularly from Italy, Hungary,
Poland, Russia, and Czechoslovakia, as well as
northern European immigrants from England,
Germany, and Scotland. Lewis (1987) finds that coal
mine operators pursued a policy of "judicious mixture"
to ensure that their labor force was composed of a
combination of African-Americans, native whites, and
foreign workers. Reducing the social force of any one
ethnic group was meant to "divide and conquer" in
order to achieve maximum control over labor and to
minimize the costs of production. In an arena in which
power was so thoroughly dominated by the coal
interests, however, workers instead forged a class
movement characterized by ethnic and racial
cooperation. The United Mine Workers of America
(UMWA), established in 1890, offered union
membership to all workers and a democratic alternative
to the organization of their work. Black miners were
equally active in the UMWA, which included
participation in the bloody strikes of the 1920's and
30's. With the mechanization of coal production from
the 1930fs to the 1950fs, however, hundred's of
thousands of miners lost their livelihoods. Black
miners were most affected by the changes and, by
1970, fewer than 4,000 remained in the industry (Lewis
1987).
Today, black Appalachians constitute approximately
8% of the region's population, and they have been more
affected by the processes of out-migration and
urbanization than white Appalachians. Blacks in the
region suffer from severe socioeconomic deprivation,
with the result that poverty is higher here than that for
blacks nationally. William Turner summarizes their
condition by saying, "The only group of persons worse
off economically than black Appalachians are rural
blacks in the United States. Rural Appalachian blacks
are likely among America's poorest people"
(1985b:257). Although little social research exists on
black communities in the region, there is evidence that
black Appalachians share a distinctive identity as
"black" and "Appalachian," and that this sets them
apart from other African-Americans in the US. Turner
states that as such, black Appalachians are "a racial
minority within a cultural minority" (1985a:xix).
The Europeans who came to mine coal from the
1890's to the 1920's also contributed to the religious
diversity in the region. Most eastern and southern
Europeans were Catholics, and priests followed them
into the mountains to establish parishes. Many of the
immigrants later migrated to northern cities with the
increasing elimination of mining jobs in the 1930's, so
the Catholic Church moved to establish missions in
mountain county seats and to conduct "radio
preaching" in order to stem the decline of Catholicism
(Wolfe 1980). Other Europeans moved into the
mountains in the 20th century to find permanent homes
for themselves and their families. Many of these
immigrants were peddlers, merchants, restauranteurs
and other small business operators, and skilled
craftsmen such as shoemakers and tailors. Small
Jewish communities grew up in larger towns and cities
in Appalachia, and occasional synagogues were
founded.
Ethnic diversification continues to affect the
Appalachian region. With the end of the Viet Nam
War, a number of Hmong communities were
established in rural areas through the missionary efforts
of local churches.
Changing agricultural labor
demands have introduced new ethnic groups. With the
growth of the Christmas tree industry in western North
Carolina, for example, Hispanic farm workers
(particularly Mexicans and Central Americans) have
come to reside in the region.
�Despite this ethnic diversity, the vast majority of
Appalachian people (more than 80%) are descendents
of northern European settlers who arrived in the region
in the 18th and early 19th centuries. Immigrants to the
ports of Pennsylvania during this period in American
history were largely Scottish, English, Irish, and
German, with small numbers of Welsh Baptists, Swiss
Protestants, and French Huguenots (Fischer 1989).
Facing conflict with the Quakers already settled in the
vicinity of Philadelphia and Newcastle, these
immigrants soon were pushed southward along the
Atlantic coast into the Carolina piedmont and to the
southwest into the Valley of Virginia along the
Shenandoah River. They were encouraged by officials
to move into the "back parts" of the colonies where
they would form a buffer between the seaboard
settlements and the Indians. The German immigrants
tended to concentrate in the northern end of the Valley
of Virginia and in Pennsylvania, becoming known as
"Pennsylvania Dutch" (a derivative of Deutsch). From
the Greater Appalachian Valley, the remaining
immigrants migrated to the headwaters of smaller
rivers in the Blue Ridge Mountains, and finally spilled
across the Appalachian divide into the Allegheny
Mountains of Kentucky and Tennessee. The region
was largely settled by 1850, and later American
immigrants began to by-pass the mountains, moving
through the Ohio River Valley into the newly acquired
western territories. No more large-scale migrations
into the Southern Highlands occurred after the midnineteeth century, until the recent influx of newcomers
connected with the tourist and recreation economy of
the late twentieth century.
The vast majority (perhaps four-fifths) of the early
European immigrants to Appalachia were Scotch-Irish
(Campbell 1921), over 250,000 of them having arrived
before the American Revolution. The Scotch-Irish
were something of a hybrid group bound together
mainly by religion and political economy (Fischer
1989). They were Calvanists, originally from the
Scottish lowlands and the north of England, who were
forced to emigrate to northern Ireland after it was
claimed by England in the early 17th century. In
Ulster, they continued to be exploited by high rents,
low wages, and heavy taxes imposed by an Anglican
elite and a Catholic majority. Joined by Irish
Protestants who were also seeking religious and
political freedom, many moved on to America. Free
land in the western territory after the American
Revolution attracted the settlers to the Appalachian
mountains, but they were also impressed with the
beauty of the land, the rich forests, the plentiful rivers
and springs, and moderate temperatures. Scotch-Irish
immigration peaked in the late 18th century and faded
by the mid-nineteenth century.
By this time,
nonetheless, the Scotch-Irish had imposed their own
distinctive cultural imprint.
In Albion's Seed (1989), historian David
Hackett Fischer argues that the ethnic heritage of these
settlers in the Appalachian region was sufficiently
distinctive to create a different American regional
culture he calls "Borderlands/Backcountry." Citing
differences in 26 folkways, including speech, family,
gender, death rituals, magic and religion, and
conceptions of order, power, and freedom, Fischer
contrasts the Borderlands/Backcountry subgroup with
three others of British origin: the Puritans of New
England, the southern English who settled in Virginia,
and the North Midland English and Welsh of the
Delaware Valley. Fischer points out that many of the
Scotch-Irish continued to migrate within the United
States, moving into Arkansas (the Ozarks) and Texas in
the nineteenth century and to Arizona and southern
California in the twentieth century.
Despite this historical pattern of migration and
concentration in the Appalachian region, the ScotchIrish as such have not maintained much of an ethnic
identity in the US. As a result, there is little consensus
on the contemporary nature of the Scotch-Irish as an
ethnic group, nor do scholars regard this group as being
relatively significant among Euro-American ethnic
groups. For example, few references are ever made to
the Scotch-Irish as a contemporary American ethnic
group in the scholarly literature. On one hand, this is
an extension of the tendency (until recently) not to
recognize Euro-American ethnic groups as such in
ethnicity texts. On the other hand, even among EuroAmerican groups, the Scotch-Irish are relatively
obscure in the literature. In a recent study, I reviewed
20 general ethnicity texts that discussed European
American ethnic groups, and found that the most
commonly cited groups included Jewish, Irish, Italian,
German, English, and Polish Americans (Keefe 1992).
Each of these groups were cited in 12 or more of the
books and were cited on at least 3% or more of the total
pages in these books. The Scotch-Irish, on the other
hand, were cited in only 8 of the books and on less than
1% of the total pages. Difficulty in establishing the
character of the contemporary Scotch-Irish is increased
by the fact that the US Bureau of the Census and most
large-scale surveys treat the Scotch-Irish not as a single
ethnic group but as an ethnic mixture of Scottish and
Irish peoples (Alba 1990:342).
�As a people, the Scotch-Irish have a weak
identification with their European heritage. Despite
being predominantly Scotch-Irish, for example,
Appalachians rarely acknowledge their national origins
(Beaver 1986), identifying more often simply as
"Americans" or as "mountain" or "country" people. In
many instances, they identify most with a state, county,
or community locale. Mary Waters (1990) notes that
the tendency to abstain from identifying with a EuroAmerican ancestry in the US is found most often
among rural Southern whites with comparatively little
education, those who might be expected, in fact, to
have Scotch-Irish background.
According to the 1990 census, only 2.3% of the US
population (5,617,713 people) claim Scotch-Irish
ancestry, compared to 13.1% claiming English, 15.6%
Irish, and 23.3% German ancestry (US Bureau of the
Census 1992). Another 5% of the US population
identify their ethnicity as "American,11 while 10.5% do
not report ancestry. There are almost as many people
who claim Scotch-Irish ancestry in California and
Texas (1,042,382) as in the Appalachian states of West
Virginia, Virginia, Kentucky, Tennessee, North
Carolina, and Georgia (1,173,584). At the same time,
almost three times as many people in these
Appalachian states as compared to California and
Texas claim "American" ancestry (3,606,051 vs.
1,539,560), and an even larger number does not report
ancestry at all (4,985,223 vs. 4,251,509): that is, 29%
of the population in these Appalachian states claim
"American" identity or eschew ethnic ancestry
altogether.
The above suggests a loss of a sense of Scotch-Irish
identity, especially in the mountainous parts of these
states. The reasons have to do, first, with the greater
political utility of class and regional consciousness as
opposed to Scotch-Irish ancestry among mountaineers
as a means of organizing for resources. Royce (1982),
Peterson (1980), and others have discussed the fact that
ethnicity is a strategy for which individuals and groups
may opt when it is advantageous in politically
competitive situations. The relative isolation and
homogeneity of the Appalachian population in the early
19th century would have reduced the need to enunciate
a distinct heritage in competition for the plentiful
resources in the region at the time. During die Civil
War, mountain residents, who typically were not slaveowners, frequently sympathized with the Union's
efforts and suffered political retaliation by southern
state governments during and following
Reconstruction. The late nineteenth century was also
a period when wealthy northeastern capitalists began
taking economic control of the land, mineral rights,
timber reserves, and labor in the region. In other
words, Appalachians were affected more by
competition with other white residents in their states
(many of whom were also Scotch-Irish) and with elite
Euro-American Protestants who for the most part were
absentee landlords. Regional and class identity were
more reliable bases of organization for competition
given these circumstances.
A second major reason for the loss of Scotch-Irish
ethnicity involves the transformation of the
Appalachian people as a result of their religious,
political, and economic experience in America. The
residents of the region were changed substantially by
these processes, becoming a new people: they were no
longer Presbyterians, nor simply subsistence farmers,
nor even citizens in control of their own resources or
destiny. These historical processes will be discussed in
the next section.
A final aspect of diversity deserving mention is the
increasing in-migration of non-Appalachians to the
southern subregion. Appalachia in the twentieth
century has been most characterized by out-migration
due to economic deprivation, especially in the coal
regions. With the rise of the tourism and recreation
industry in southern Appalachia in the 1970's, many
areas began to grow in population, because of the
growing number of jobs. Some counties in western
North Carolina, for example, such as Henderson and
Watauga Counties, have increased in population by 2040% since 1970.
In-migrants have often been Appalachian outmigrants returning home to begin again. More often,
however, the in-migrants are newcomers attracted by
the region's beautiful landscapes, rural lifestyle, and
relative absence of urban problems such as crime,
pollution, drugs, and traffic jams. Non-Appalachian
newcomers hail from all parts of the country, but most
often they come from nearby states or adjacent regions.
In western North Carolina, for example, they are most
frequently either from the Carolina piedmont or
Florida. Some of the new arrivals, the "back to the
landers," are urban ex-patriots seeking a simpler, less
materialistic life (Beaver 1986). Others are wealthy
second-home buyers, some of whom eventually
become permanent mountain residents.
Given that they are generally white southerners in
origin, newcomers often share a similar northern
European heritage with mountaineers. Socially and
culturally, however, they tend to be distinctive, as they
�neighbors within a common geographic location (along
a hollow or a creek) and joined together by a shared
history and moral code (Beaver 1986; Pearsall 1966).
Schools, country stores, post offices, and churches, the
typical anchors of rural community life, were
uncommon until after 1890. Mountain communities
were egalitarian and unstratified for the most part, each
adult male being essentially a subsistence farmer. As
in any rural agrarian society, economic exchange was
based on reciprocity, or the exchange of goods and
services between equals. Reciprocal exchange was
epitomized by such traditional communal farming
activities as barn-raisings, com shuckings, and quilting
parties.
are usually well-educated middle or upper middle class
whites with urban backgrounds and more mainstream
American lifestyles (Keefe, Reck, & Reck 1989).
More and more frequently, they have come into
conflict as a group with native Appalachians,
particularly over land-use issues, such as zoning, and
the legislation of moral codes, such as the sale of
alcohol (Keefe 1994). As their numbers and political
power grow, non-Appalachian residents will
increasingly become a force to be reckoned with in
certain parts of the Appalachian region. And as a
consequence of their presence, native Appalachian
residents can be expected increasingly to identify as an
"ethnic group" in the region in order to try to secure
their status in the new social order.
The fundamental basis for membership in mountain
society was kinship. This provided an idiom for
conceiving social relationships, the concept of
"equality" ultimately being based on the recognition of
common blood heritage. As a real social system,
kinship was the basis for most interaction as
mountaineers lived and worked primarily within the
extended family. In the absence of local churches and
ministers, even religion tended to be family-based, fed
by brief contact with circuit-riding preachers.
An Appalachian Ethnic Group?
While the region of Central and Southern Appalachia
is racially and culturally diverse, it is nonetheless
possible to speak of Appalachians as an ethnic group,
or perhaps more accurately an "emerging ethnic entity"
(Peterson 1980). The historical roots of this sense of
identity are complex, yet the tendency to identify as
"Appalachian" is manifested in groups as diverse as the
Cherokee in western North Carolina, blacks in the West
Virginia coal fields, and white mountaineers in rural
eastern Tennessee. Because whites predominate in the
region, the following narrative focuses on them.
A series of religious revival movements, known as
the Great Awakening, swept through the American
colonies in the eighteenth century, and the Appalachian
mountains and the rest of the South felt the impact of
the second of these movements from about 1790 to
1810. The movement emphasized revivalism which
involved frontier camp meetings, outdoor religious
services, and emotional forms of preaching. The
structure of the worship service permitted greater
participation by the congregation than was usual in the
established religions, and revivalism emphasized the
individual religious experience rather than the religious
doctrines of a particular church. The effect of the
revival movement was to convert large numbers of
people away from the orthodox religions.
Prior to 1880, Appalachia consisted of small family
farms that were largely self-sustaining, producing their
own handicrafts and relying on only a handful of
market goods (i.e. coffee, sugar, salt) that they could
not produce themselves. The Scotch-Irish came to
America with a rural culture based on a non-intensive
form of farming called "forest farming." This is
practiced in marginal areas with poor soils where it is
necessary to rotate cultivated plots frequently with
forested areas (Blethen 1994). Livestock (hogs, cattle,
and sheep) were also raised and allowed to graze freely
on the acorn and chestnut mast on the forest floor.
Hunting, fishing, and foraging wild plant foods (herbs,
berries, nuts, and greens) were important economic
pursuits. Original land grants in the mountains were
relatively small, a few hundred acres at most, certainly
a great deal smaller than those granted in the plantation
region of the Deep South, and inheritance practices
subdivided these family plots with each generation.
The movement appealed to the Scotch-Irish and
others who had been at odds with both the Church of
England and political authoritarianism. It formed the
basis of later evangelical Protestantism and the belief
that individuals might be "born again" and, through
faith, find everlasting salvation. At the heart of the
religious movement was the personal conversion
experience, the idea that every individual could
personally find God and be saved. This was an
optimistic frontier religion in contrast to traditional
Calvinist beliefs in "divine election" and predestination.
The new religious movement focused attention on the
Nineteenth-century mountain communities were not
the town-square New England variety, but rather
formed "open-country neighborhoods" made up of
intricate dispersed networks of family, friends, and
8
�individual, the adherence to a personal code of moral
ethics (including prohibitions against smoking,
drinking, swearing, and gambling), and the expectation
that people will respond vocally and emotionally when
Jesus enters their hearts.
Services became
characterized by emotional expressiveness, as male and
female participants responded to sermons with singing,
shouting, and testifying and ended services with warm
and friendly hugs, handshakes, and kisses (Dorgan
1987). The emphasis on a "literal" interpretation of the
Bible meant that formal theological training (as
demanded by the Presbyterian order, for example) was
perceived as less important than receiving a "spiritual
calling" to the word of God. So mountaineers
welcomed the less educated farmer-preacher and
emphasized religious "ordinances" following Christ's
teachings and practices (such as creek baptisms and
footwashing, an extension of the Lord's Supper) as
opposed to religious "sacraments" emerging from
formal church traditions (such as holy orders and
church-sanctioned marriage).
A result of the Great Awakening was an increased
religiosity among mountain people combined with a
decline in affiliation with national church hierarchies.
Today, the United Presbyterian Church as well as other
mainline churches have little strength in rural parts of
the region, which are dominated by small, family-based
churches, usually Baptist, Methodist, or HolinessPentecostal. Church members' sentiments were not
universally favorable to revivalism or related practices,
such as Sunday school classes, causing many divided
congregations and denominations. "Old Regular"
Baptists, a highly traditional subdenomination largely
located in Central Appalachia, for example, grew out of
the contingent of Baptists rejecting revivalism (Dorgan
1989). In Appalachia, these divisions resulted in the
emergence of numerous sectarian and independent
churches, unaffiliated with national or "mainline"
churches such as the Southern Baptist Convention or
the United Methodist Church (although Appalachian
churches were sometimes drawn together into loose
"associations"). Humphrey (1984), for example, counts
47 varieties of Baptists in the mountains, 33 varieties of
Methodists, and 18 varieties of Presbyterians.
Mountain counties today are characterized by large
numbers of simple, unassuming churches, typically
having less than one-hundred members, scattered
throughout the countryside. In a study of one rural
county with a population of less than 35,000 in western
North Carolina, for instance, 125 active churches were
identified (Keefe et al. 1985).
The Great Awakening, then, served to loosen white
southern highlanders* religious ties to their European
homeland. Black Appalachians, like other black
southerners, were also transformed by this spiritual
movement and their religious practices and
organization also reflect a preference for emotional
rituals, revivalism, and sectarianism (Williams 1982).
Fifty years later the Civil War served to magnify the
differences between the yeoman mountain farmers and
the plantation aristocracy of the lowland South,
furthering regionalism in Appalachia. As previously
mentioned, slavery was less important in the mountains
where the vast majority of whites were small
landholders without the need or wherewithal to buy
slaves. Throughout the nineteenth century, the
mountainous parts of the southern states were often at
odds with flatland rivals in statewide political battles in
which the politically-dominant plantation owners
generally were victorious. Economic ties of the
mountain mercantile class (those most likely to be
slaveowners) generally were with the South once the
War Between the States broke out (Inscoe 1989), but
sentiments varied among other mountaineers, many of
whom joined the Union at the same time that others
joined the Confederate forces. In addition, there were
the "Tories," those with split loyalties who avoided
enlisting in either army and who fled to the hills as
"outliers" when they were in danger of being
conscripted (Blackmun 1977). In certain mountain
counties, the majority of sentiment was Unionist, and
following the war these counties became Republican
partisan strongholds, further distancing themselves
from the Democratic majority in postbellum southern
states (McKinney 1986).
These divided loyalties during the Civil War
politically severed mountaineers from their southern
compatriots. Hill people became stereotyped in state
legislatures as backwards and undeserving of
citizenship and equal rights, thus justifying the states'
neglect of mountain roads and schools. Severe
socioeconomic deterioration in the mountains was the
result in the late nineteenth century.
The industrial development of Appalachia beginning
around 1880 produced further radical transformations.
Railroads began to be constructed with federal
government subsidies, and by 1930 they criss-crossed
the entire region leaving only a few counties without
access (Eller 1982). The railroads were vital to the
extraction of Appalachia's natural resources, especially
timber and minerals, such as mica, iron ore, and coal,
which had to be hauled to cities and markets where
secondary industries converted them into goods and
�where consumers were available to buy them. Mining
and timber companies began purchasing land and
mineral rights, so that by 1900 many mountain counties
had seen outside capitalists buy up a majority of these
natural resources. Speculators often used unethical
practices to buy and sell land. They also used such
mechanisms as the "broad form" deed which made it
possible, later, for the owners of mineral rights to
engage in strip mining and, to destroy land that was
owned by local mountaineers. The inevitable rising
taxes that accompanied industrialization and the need
for money in an increasingly market-driven rather than
subsistence-based economy forced farmers to sell out
or at least to take on wage labor to supplement their
household incomes. The federal government also
began to condemn and buy land as the National Park
system (the backbone of modern tourism in
Appalachia) took shape in the early twentieth century.
By 1920, two brief generations following the
penetration of capitalism, only 20% of the labor force
in Appalachia was practicing fulltime farming.
modernization and industrialization in Appalachia did
not result in a rising standard of living but instead
produced one of the highest rates of poverty in the
country.
Many of the problems that Appalachia experiences
in catching up economically with the rest of the nation
today are rooted in this early industrial period when the
ownership of basic resources passed from local to
absentee ownership and the region became an "internal
colony" controlled and exploited by outside interests
(Lewis etal. 1978). The Appalachian Land Ownership
Task Force (1983) found that in 1980 three-quarters of
the surface land rights and four-fifths of the mineral
rights in Appalachia are in the hands of absentee
owners, mostly large, private (and multinational)
corporations as well as federal government agencies
(such as the Park Service and Forest Service). Local
governments have inadequate monies for local
development because federal lands are tax-exempt
while private corporations benefit from lucrative tax
breaks. Moreover, the extractive industries of the
region do not promote the development of an economic
infrastructure of roads, health clinics, good schools,
and public services that would make the region
attractive to a more diverse industrial base.
Appalachia is often portrayed as a premodern,
nonindustrial region, but in fact the area has long been
industrialized, and indeed it provided much of the raw
materials fueling American economic development
during the twentieth century.
What does set
Appalachia apart from the rest of the country, on the
other hand, is the character of its economic
development as a peripheral region providing labor and
natural resources to the core of the country, which
skims off the profits by marketing those resources
(Walls 1978). Appalachia lacked an intermediate stage
of commercial transformation in the nineteenth century
(Pudup 1990). This was due to a number of reasons,
including a relatively homogeneous subsistence
economy based on small, self-sufficient household
farms with few transportation or production ties to
external markets. When outside markets began to
demand the timber and coal that the region contained,
the capital required to exploit those resources had to
come from outside investors also. And so, unlike the
northeast where a slow evolution of capitalism and
indigenous commercialization promoted the emergence
of a substantial middle class and a local base of capital,
Appalachia was "invaded by mature capitalist
institutions11 which converted the local population
rather suddenly and almost wholly into working class
laborers (Billings, Blee, & Swanson 1986). Moreover,
the nature of the industries that took root in the region
did not require a sophisticated labor force, and
management actually benefitted from the perpetuation
of a monoeconomy which forced workers to accept
undesirable jobs for low pay. As a consequence,
The historical experience of Appalachians has served
to make a heterogeneous racial and cultural aggregate
of people into a regional group with a shared identity.
Caught up in the Great Awakening, Appalachians, like
many other Americans, transformed their orientation
from the Old World to the New World.
Presbyterianism and Scotch Irish ethnicity lost strength
as southern highlanders adapted to new cultural and
environmental situations, in which regional residence
and social class membership became important
determinants of status in society. Links of political and
economic geography to the rest of the South were
fractured by the Civil War, so that Appalachian areas
became, and remain, the "untended backyards" of
southern states. Perhaps most importantly, Appalachia
was transformed in a few brief decades around the turn
of the century from a Jeffersonian frontier society to a
highly stratified one in which the vast majority of
native mountain residents became exploited laborers in
a peripheralized region. These processes set the stage
for an emerging consciousness among Appalachians as
a people with a single historical experience, a shared
relationship to the outside world, and a common
destiny. Appalachian ethnic awareness was promoted
by the images developed as others came to confront
mountaineers at home and outside the hills.
10
�The Construction of Cultural "Otherness"
In another post-Civil War development, Protestant
missionaries sent by northern churches entered the
mountains as part of an aggressive evangelization and
Americanization of the South. As Shapiro points out,
"The local-color writers saw Appalachia as an
'unknown1 land because they had never been there. In
the same way, the home missionaries of the northern
Protestant churches saw Appalachia as an 'unchurched1
land because their denominations were not represented
there" (1978:32). By the mid-1880's, all northern
Protestant churches (including Presbyterian,
Episcopalian, Methodist, Congregational, and Quaker)
had missionaries working among southern whites. In
addition to ministerial practice, benevolent work in the
southern mountains primarily consisted of establishing
schools and churches. These were the institutions
perceived both as absent and as critical in addressing
the social as well as geographic "isolation" of mountain
people caught in "Rip Van Winkle sleep". Mary
Noailles Murfree's In the Tennessee Mountains was
used as a text to inform new missionaries about the
region. Going beyond the local-color definition of
mountain life as distinctive and degenerate, however,
denominational workers found in Appalachia a social
problem and the need for uplifting a people not so
much degraded as "not yet graded up." The
establishment of mission churches, mission schools,
and colleges institutionalized the acceptance of
Appalachian otherness.
It is perhaps no coincidence that at the same time
industrialization and peripheralization of the
Appalachian region was taking place in the late
nineteenth century, an image of Appalachia as a
"strange land and peculiar people" began to emerge in
the American consciousness. Henry Shapiro (1978:x)
argues that "the idea that the mountainous portions of
eight or nine southern states form a coherent region
inhabited by an homogeneous population possessing a
uniform culture" was an invention of those from
outside the region who came to perceive Appalachians
as deviant from the American norm. The peculiarities
identified as "Appalachian" contributed to their further
estrangement from other white Protestants, other
southerners, and other agrarian peoples in the nation.
The growing perception of Appalachian cultural
deviance served in many ways to justify the process of
dispossession of their land and resources during the
new era of capitalism. Moreover, the perception of an
Appalachian "problem" fueled literary, missionary, and
scholarly work, the popular media, and governmental
policy for the next century.
Appalachia was "discovered" by the nation in the late
nineteenth century on the pages of American
magazines and books of fiction. A literary period now
known as the "local-color movement" flourished
following the divisive Civil War. Describing littleknown places and people in picturesque styles, often as
travel sketches and short stories laced with dialect,
local color writers provided middle-class readers of
Scribner's, The Atlantic Monthly, and Harper's
Magazine with a vision of national unity through
symbolic contrast with the "other." Appalachia and
other remote areas of the country were exploited for
their geographic wonders, biological curiosities, and
cultural peculiarities thought to be of interest to readers
in eastern cities. Fiction writers, such as Mary Noailles
Murfree (In the Tennessee Mountains 1884) and John
Fox, Jr. fThe Little Shepherd of Kingdom Come 1903;
The Trail of the Lonesome Pine 1908), dealt with
dramatic "confrontations between characters
emblematic of the two cultures—feudists and lawmen,
moonshiners and revenue officers, dirt-farmers and
mining engineers, plain folks and politicians, mountain
girls and city boys" (Shapiro 1978:20). The earlier,
romantic notion of Daniel Boone-style pioneers of
mountain stock was replaced by unappealing images of
Appalachians as anarchists, criminals, rubes,
degenerates, and victims. These stereotypic images
were to be reinvigorated by the popular media of the
twentieth century.
Berea College in Berea, Kentucky, was founded in
1855 as part of the missionary movement in the
mountains. One of the most influential presidents of
Berea College, William Goodell Frost, was the first to
name the region, known in the nineteenth century only
as "the central South." In an essay written for Atlantic
Monthly in 1899 entitled "Our Contemporary
Ancestors in the Southern Mountains," Frost revealed
the architecture of his vision of the region he called
"Appalachian America," a region of mountainous
terrain in the heart of the most civilized area of the
nation, a region in which a hardy race of pioneers
continued to live in traditional ways. By naming it,
Frost gave Appalachia a sense of coherence and
homogeneity not otherwise apparent (Shapiro 1978).
By referring to mountaineers as the nation's
"ancestors," Frost conceded Appalachians
backwardness while at the same time urging national
concern about the people who were, he argued, close
kin of mainstream whites.
Benevolent workers were attracted to the southern
mountains in the hopes of providing both their version
of the gospel and a sense of community which was
11
�mountaineers through the "perpetuation" of local
traditions. They, too, however, more often resulted in
cultural invention and manipulation. The most wellknown folk school, the still extant John C. Campbell
Folk School in Brasstown, North Carolina, for
example, was founded on the Danish School model and
even incorporated Danish folksongs and dances into the
curriculum. David Whisnant (1983) points out that
these intentional and systematic cultural interventions
were part of a "politics of culture" in which nonAppalachians sought to shape the thoughts and values
of mountaineers, with profound consequences for their
identity and sense of cultural differentness.
thought to be necessary for social progress. The
perceived "absence of community" in the mountains
was believed to be a consequence of isolation. The
outbreak of feuds around the turn of the century,
especially in Kentucky, West Virginia, and Tennessee,
confirmed emerging notions that Appalachians were
stubbornly independent, anti-social, and lawless.
Journalists' descriptions of family feuds and related
murder trials were eagerly consumed by readers of
eastern newspapers. Some social historians assert that
feuding was customary among mountaineers, who
brought a tradition of autonomous family retribution
for wrongdoing from northern Britain (Fischer 1989).
In a careful investigation of the famous Hatfield and
McCoy feud, however, Altina Waller (1988) finds a
better explanation in the invasion of the mountains by
industrial capitalism and the development of economic
inequality in what had been a kin-based, egalitarian
culture. "Devil Anse" Hatfield was, in fact, an
ambitious local entrepreneur who in 1877 started a
timber company employing 30 men, including many
family members; while "ol1 Ranel" McCoy and his
sons, who were landless, had a more marginal standing.
The twelve year feud beginning in 1878 engaged old
family animosities in a changing economic climate
where competitors were not only outside vested
interests but indigenous market capitalists.
Coincident with the founding of folk schools in the
mountains was the labeling of Appalachians as "folk,"
or a separate people with a "folk culture." In one sense,
this was meant to justify the establishment of folk
schools which assumed the pre-existance of a crafts
tradition which might be preserved through the school
curriculum. It also lent a sense of the "primitive" to the
products produced and marketed by mountain men and
women, such as quilts, knotted bed canopy fringes,
coverlets, rag carpets, oak-splint baskets, brooms, splitbottomed chairs, wooden toys, and handmade
dulcimers and banjos. The handicraft movement in
Britain and America at the turn of the century brought
recognition to the aesthetic merit of "naif1 work and
helped to foster an economic market for handmade
crafts. Folk school directors were quick to exploit the
commercial advantages for mountain people who were
promoted as "the conservators of our common AngloSaxon heritage, preservers of the folk culture of Merrie
Olde England and the individualism of the days of the
pioneers" (Shapiro 1978:218).
The settlement movement came to Appalachia at the
turn of the century to bring social and cultural
development to a people perceived as lacking both.
Part of the larger settlement house movement (whose
best known example is Jane Addams1 Hull House in
Chicago), various rural settlements were established in
the southern mountains by young women, graduates of
northern colleges such as Smith and Bryn Mawr, who
were called to the Victorian ideal of "social
guardianship" (Whisnant 1983). Settlement schools
were meant to "revive" local culture, especially
handicraft traditions, among the adult population, but
the crafts also reflected outside influences brought in
by the outsiders themselves, including new weaving
patterns and woodworking styles (Whisnant 1983).
Settlement school directors also imported and taught
their own version of proper cooking, housekeeping,
and even silverware table settings, as well as new ways
of celebrating holidays such as Christmas, and new
forms of socializing on Saturday nights (foregoing
moonshine drinking and raucous dancing for more
genteel rituals).
Efforts to preserve traditional folksongs and
folkdances followed in the footsteps of the crafts
"revival" movement. Olive Dame Campbell, the wife
of John C. Campbell, was one of the early folk ballad
collectors in the mountains. At her urging, Cecil J.
Sharp, a well-known British collector of folksongs,
arrived in 1916 and worked with singers in North
Carolina and Tennessee. Sharp was impressed with the
quality of his informants whom he characterized as
"just English peasant folk [who] do not seem to me to
have taken on any distinctive American traits. They
talk English, sing English, behave English!" (quoted in
Whisnant 1983:116). The collaborative efforts of
Campbell and Sharp produced a joint publication,
English Folk Songs of the Southern Appalachians
(1917). Like other Appalachian collections published
in the period, special emphasis was put on folksong
"survivals," including ballads such as "Barbara Allen,"
Folk schools, introduced at the same time as rural
settlement schools, were more specifically intended to
provide self-sustaining economic alternatives for
12
�and often negative connotations of the hayseed
trappings used to characterize mountain culture.
"Pretty Polly," "Young Edward," and "Lord Daniel's
Wife." This enchantment with the persistence of
English folk culture was echoed in the delight of
settlement school directors who discovered
"Elizabethan" speech patterns in the mountains. At
Hindman Settlement School in eastern Kentucky, for
example, children were even taught Shakespearean
plays in order to employ their supposedly natural
"Elizabethan" dialect, with less than satisfactory results
(Whisnant 1983).
The "hillbilly" stereotype emerging in the music
industry has been echoed in other commercial media
throughout the twentieth century. Cartoonists, such as
Paul Webb drawing for Esquire (1935-1948), Al Capp
who created "Lil Abner" in 1934, and Fred Lasswell
who draws "Snuffy Smith," have used mountain
images to communicate themes of ignorance, poverty,
sloth, filth, immorality, and regressive attitudes and
behaviors. Jerry Williamson (1995) demonstrates the
significance of Appalachian themes in Hollywood
productions beginning with early silent films to the Ma
and Pa Kettle films of the 1940's and the popular
"Beverly Hillbillies" television show of the 1960's, all
of which projected familiar stereotypes both to the
nation and to foreign audiences. Commoditization of
these Appalachian images continues in the expanding
tourism industry in the region, where "hillbilly" trading
posts and "country" emporiums sell "outhouse" and
moonshine-still replicas, 'coon dog portraits, and
"chewing tobacco" chewing gum.
Where English survivals were absent, settlement
schools sometimes introduced them. One of the most
arresting examples of this involves morris dancing,
believed to be a form of "ancient pan-European
seasonal pagan observances associated in some way
with fertility" (Whisnant 1983). Disturbed by his
failure to find morris or related sword-dance
ceremonies in the southern mountains, Cecil Sharp
began to teach them to mountain people at Hindman
and Pine Mountain Settlement Schools in Kentucky
immediately following his arrival. When mountain
folk festivals celebrating traditional music became
popular in the 1930's, such as White Top Folk Festival
in Virginia, they served as venues for the further
preservation of these "reintroduced" dance forms.
Since the late nineteen-hundreds, Appalachia has
attracted considerable national attention. Images of
Appalachian people and culture have been created and
manipulated by those from outside the region to serve
as examples of "otherness," from visions of hardy selfsufficient pioneers to isolated and lawless anarchists to
romantic reminders of America's ethnic folk heritage to
laughable rubes hopelessly out of place in modern
America to a national social problem that must be
solved. This "invention" of Appalachia, as Allen
Batteau (1990) points out, is perhaps best understood as
a consequence of various national concerns, including
unification, modernization, and social progress. The
images have had an impact on Appalachian peoples1
identity, forcing them to deal with the pervasive
negative stereotypes. But the images have little to offer
in providing an accurate understanding of Appalachian
culture and values. For this, we must turn to the
ethnographic evidence and to community studies
carried out in the region,
Mountain music was reshaped even more
dramatically for a mainstream American market with
the influence of the commercial music industry. As
early as 1923, mountain musicians were producing
phonograph recordings with impressive commercial
success, and the record industry adopted the name of an
early group from Galax, Virginia, the "Hill Billies," to
refer to the entire musical genre (Tribe 1990). The
most significant musical group to emerge from the
southern mountains during this early period, the Carter
Family, included in their repertoire all of the diverse
traditions in Appalachian songmaking: British and
traditional American ballads, nineteenth century
popular and Victorian songs, sacred lyrics and gospel
music, and African American and minstrel show songs.
Later artists from the mountains were encouraged to
stray from these traditions and to write their own songs
to appeal to a wider commercial market, and new
musical styles were created, such as Bluegrass music in
the 1940's which modernized "old-time" string band
music. Radio shows, such as the Grand Olf Opry,
popularized music that became known as "country and
western," in which the country or mountain music
oftentimes had to adopt a semi-comic, dialect-peppered
style in order to gain acceptance. Television versions
of these shows, such as "Hee Haw" (which enjoys great
popularity in the rural south), perpetuate the humorous
Appalachian Culture and Values
Ethnographers documenting Appalachian culture
agree on a number of core values and cultural features.
While most ethnographic studies of Appalachian
communities concentrate on rural areas and very small
towns, similar cultural traits are found in studies of
Appalachian migrants to cities outside the region. It
must be assumed that these traits are also characteristic
13
�Beaver 1986; Halperin 1990). Economic strategies in
mountain communities are anchored on the family
household and the value of adequacy (Halperin 1990).
The goal is to make ends meet and provide for family
needs, not to make profits. Halperin (1990) describes
the "multiple livelihood strategies" used by rural
families to adapt to the local mix of capitalist and
precapitalist economies where reliance on a single
livelihood would be risky. By using their land to plant
subsistence gardens, working in factories as temporary
wage labor, doing their own household repairs,
bartering with neighbors and friends, patronizing flea
markets where cheaper factory seconds and used goods
are sold, and relying on kin networks to pool resources,
mountain people resist dependency on capitalism and
are able to maintain a traditional rural lifestyle. This
lifestyle also reinforces homogeneity in the social
community where no one is a specialist and everyone
is relatively autonomous. And as Halperin (1990)
points out, while this livelihood may result in low
income, people do not think of their family as "poor."
Instead, their ability to "make do" is held in high
esteem within the local community (Stephenson 1968).
of Appalachian residents in larger towns and cities
within the region, although this awaits confirmation by
future urban ethnographies.
Appalachian culture is not wholly unique, for it
shares much with rural Southern culture and rural
agrarian lifestyle in general, but as a pattern of ideas
and expectations for behavior Appalachian culture is
coherent and distinctive. The following attributes are
essential to Appalachian culture: egalitarianism,
independence and individualism, personalism, an
avoidance of conflict, familism, a religious world view,
and a sense of place.
Egalitarianism
Egalitarianism, of course, is an American cultural
value. As Beaver (1986) explains, however, in
Appalachia it has more significant "mythical" qualities
as a guide for people's actions. Certainly mountain
communities are stratified (although given the rural
economy, status differences may not be as great nor as
pervasive as in urban areas). But great effort is taken
to ignore these differences in social relations. Common
in mountain communities is the country store or small
cafe where all members of the community, rich and
poor, meet and greet one another by first names and sit
down together at the same table or counter. Everyone
expects to be treated with identical courtesy and to ask
about each other's families. There is little pretense or
status differentiation by dress in mountain
communities, where everyone dresses informally and
women wear little or no make-up (Hicks 1976).
Conspicuous consumption is avoided. People in
general are self-deprecating, often by joking about
themselves (Miles 1975). Loyal Jones (1994) refers to
this as the value of humility. Those who set themselves
above others are referred to pejoratively as "getting
above their raisin'," "uppity," or "big shots." Even
community leaders seek to be self-effacing in the
public eye. Mountaineers identify with "Jack" of the
Jack-in-the-Beanstalk folk tales, where Jack through
courage and intellect is able to outwit the giant and the
fancy folk. Also like Jack, mountaineers may engage
in petty theft or vandalism as sanctions against those in
the community with condescending attitudes or
pretensions of higher class (Beaver 1986).
As the preferred form of economic exchange,
reciprocity with family, friends, and neighbors also
reinforces egalitarianism. Miles (1975) was struck by
the extensive hospitality mountain people bestowed on
visitors when she lived in eastern Tennessee during the
late nineteenth century. Neighborliness is still highly
valued and people are expected to repay "kindliness for
kindliness" (Miles 1975). Mountain people like to
come visiting with gifts of food or produce from their
garden. The most common form of gathering is the
"covered dish" dinner where all participants come
together to share homemade foods. The food is
plentiful but simple fare found on common tables and
rooted in the people's subsistence heritage: plates of
fried chicken, country ham biscuits, and country style
steak; pots of chicken and dumplings and trays of
deviled eggs; vegetable dishes such as green beans,
pinto beans, corn on the cob, cole slaw, green salads,
sliced cucumbers, and vine-ripened tomatoes; fresh
fruit and fruit cobblers; cornbreads and other
homemade varieties of bread; pies, cakes, and banana
puddings; together with jugs of sweetened iced tea and
Kool-Aid as well as pots of hot coffee. Music (also
homemade) is often provided at these events and
reinvokes the themes of simplicity and family, as
individuals are taught to sing and play traditional pieces
by close relatives and family members often come
together to form musical groups (a pattern also found
in professional Country music groups such as the
Carter Family and the Statler Brothers).
One idiom for expressing this assumption of
sameness is kinship. As Bryant (1981) points out, the
commonly heard "We're all kin here" indicates that
people in the mountains perceive themselves to be like
one another, of the same flesh and blood. And non-kin
are often incorporated into kinship networks as equals
through the extension of kinship terms and obligations
14
�1976). This expectation of "unbounded personal
freedom" echoes an ancestral dislike of
authoritarianism and a tendency to "take care of our
own" (Fischer 1989). It is sometimes expressed in a
preference for solitude, as in the common pastimes of
fishing and deer-stand hunting.
Individualism and Independence
Like egalitarianism, individualism and independence
are core American values that have special significance
and distinctive manifestations for Appalachians.
Whereas mainstream Americans interpret individualism
as the right to non-conformity and equal opportunity to
compete for the American Dream, Appalachians
emphasize the aspects of sovereignty and self-reliance.
This is captured in a story related by Loyal Jones:
Sovereignty is sought in social as well as personal
life. Mountaineers have a tradition of resisting
governmental authority, as illustrated in the region's
moonshining heritage. Using techniques brought from
the British Isles, mountaineer entrepreneurs capitalized
on Prohibition, turning low-profit corn into higherprofit corn liquor and running it to major cities outside
the region, for the most part avoiding federal law until
its production became less profitable in the 1950's.
Halperin (1990) also finds widespread evidence of
resistance in traits such as keeping the homeplace debtfree (and therefore beyond outside control) and
omission of cash odd-jobs from income taxes. Beaver
sees this mountain value of autonomy more generally:
Several years ago there came a great
snowfall in western North Carolina.
The Red Cross came to help people
who might be stranded without food
or fuel. Two workers heard of an
old lady way back in the mountains
living alone, and they went to see
about her, in a four-wheel drive
vehicle. After an arduous trip they
finally skidded down into her cove,
got out and knocked on the door.
When she appeared, one of the men
said,
"Howdy, ma'am, we're from
the Red Cross," but before he could
say anything else, the old lady
replied,
"Well, I don't believe I'm agoin to be able to help you any this
year. It's been a right hard winter."
(Jones 1994:51)
Independence, then, means "minding
one's own business" or not meddling
in other people's business and, at the
same time, expecting to be left alone
in managing one's own affairs. This
concept is expressed most frequently
in the simple phrase "Ain't nobody
gonna tell me what to do." (Beaver
1986:153)
Appalachian individualism may be described as
"cooperative independence" whereby people cooperate
in order to preserve their autonomy and not as a form
of anarchism which threatens social harmony
(Stephenson and Greer 1983).
The self-sufficient aspect of individualism is that
which is admired in the pioneer past and glorified in
Foxfire's Aunt Arie (Wigginton and Benett 1984).
Social workers find that mountain people resist charity
and social welfare programs despite possible need.
Beaver (1986) finds the qualities of common sense,
dependability, cooperation and, especially, hard work
are the basis for an individual's sense of "worth" and
respect in the community. Not working or living on
public assistance or transgressing moral codes may
result in being considered "worthless" and exclusion
from the social community.
Personalism
Like Southerners in general, Appalachians appreciate
and encourage personal face-to-face relations.
Mountain people interact with others based on a
recognition of their individuality. This is reflected in
a story told by Helen Lewis in which she asks a young
boy named Johnny what he will be when he grows up,
and he says "I'll be Johnny" (Lewis et. al 1978). Hicks
(1976) remarks that personal status, based on family
background, reputation, leisure pursuits, indebtedness,
and personality, is something everyone has, like class
status, although personal status may be more important!
In many mountain counties, this personal status is
reflected in the tendency to name roads after
individuals. In Watauga County, NC, for example,
33% of the rural roads incorporate personal names,
Children as young as five years old in rural
Appalachia learn to take on tasks in the household and
contribute to the family's livelihood (Beaver 1986).
They are given increasing responsibility with age,
particularly in farming families. Young boys are
encouraged to use guns and begin to hunt by the time
they are ten or twelve years old. Children, particularly
boys, are allowed to be venturesome and are expected
to take care of themselves with little supervision (Hicks
15
�Avoidance of Conflict
Like Southerners in general, Appalachians tend to
promote peaceful cooperation and avoid conflict in
personal relations. To keep the peace, they do not
make public pronouncements about someone or
something they dislike. These opinions may be
expressed privately, however, or as part of gossip
controls. In fact, the common way an individual learns
she or he has done something unacceptable is "through
the grapevine" and not in a face-to-face encounter.
This avoidance of hostile expression is learned at an
early age. Preschoolers are encouraged to share and
play in harmony rather than speak in anger (Kovarski
and Braswell 1994). Aggressive, even authoritative,
behavior is considered rude outside the family, and
leadership roles are often rejected due to a concern that
they might involve uncomfortable confrontations
(Hicks 1976; Stephenson 1968). Successful leaders are
able to avoid arguments and bring about consensus
without offending others.
such as "Preacher Billings Road," "Moretz Family
Farm Road," and "Julia Pearson Lane."
Mountain people are well-known for their
friendliness and hospitality toward others. Weller
(1965), for example, recounts a story about Peace
Corps trainees from the University of Kentucky who
were sent without money to hollers in eastern Kentucky
and, by relying on the neighborliness of mountain
people, were able to find food and shelter for a week
without any trouble.
Mountaineers make eye contact with people when
interacting, transforming even mundane economic
transactions into personal ones. A local student of
mine working at a supermarket in Boone, NC,
remarked that she refused to begin ringing up each
customer's purchases until the customer looked at her
when she said "Hello." Social contact is also evident
on the roadways, as drivers frequently honk, wave, or
raise their hand or index finger in recognition of other
drivers or pedestrians. Socializing in general is
extremely important and marks all activities taking
place in the community. Sherrod (1991) finds
"fellowship" even more significant than ministerial
activities among congregations in mountain churches.
Churches typically have Wednesday night fellowship
suppers, Homecoming Day honoring those who have
moved away, and "dinners on the ground" in the
summertime. Funeral services include a "visitation"
the day before when family and friends come together,
often in the home of the surviving relatives, to share
personal support and intimate time together.
Social mechanisms to avoid conflict include the use
of formal etiquette, giving indirect advice, couching
reprimands with humor and jokes, and exhibiting a
taciturn demeanor. There is a high regard for manners
and etiquette as a means of avoiding personal offense.
Mountaineers address others as "Sir" or "Mfam" not to
indicate high status but as a form of respect. Mountain
people avoid giving advice directly to non-kin,
preferring to introduce the topic indirectly by saying
something like: "Now, if it were me,...." After a year
of rancor and dissention in the high school in my
county, a well-respected member of the community
who had recently been stricken by cancer wrote a letter
to the editor lamenting that he had not done more to
prevent the school's problems, by implication asking
what excuse others had for not taking action. George
Hicks gives a good example of the way in which selfdeprecating humor is often used to defuse situations
and reinforce an "ethic of neutrality:"
Honest and fair dealings make for strong personal
ties among mountaineers, who are loyal to those they
trust (Jones 1994). Verbal agreements are binding and
do not require paper authorization.
While
mountaineers count on each other to keep their word,
there is a general mistrust of government, institutions,
and the rich (Stephenson 1968; Weller 1965). People
work hard to be likable and accepted, and there are
strong sanctions in the community if they stray. As
Beaver notes, "gossip is in perpetual movement along
complex and overlapping networks.... [It] provides the
medium for expressing personal and collective
evaluations of behavior and allows one to avoid faceto-face encounters" (Beaver 1986:162). For these
reasons, privacy is keenly sought and protected.
Families guard secrets in order to stay out of the gossip
chains, often making it difficult for social workers,
counselors, and other professionals to work openly
with clients (Fiene 1994).
In a long discussion of county
politics one afternoon,... Carmon
Mitchell attempted to make a joke at
his friend's expense by saying,
"Ollis, why don't you run for jackass
of Kent County?" Ollis puffed his
pipe silently while the other men
roared. Sensing that he had gone too
far, Carmon added, "111 just step
down and let you take the office."
(Hicks 1976:90)
16
�than the mainstream American family: (1) the
Appalachian family includes not just the nuclear family
but an extended "family group" consisting of parents,
their grown children, and their families, (2) individuals
tend not to act alone but always to orient their behavior
with reference to the family group, and (3) blood
kinship is the strongest basis for social ties, superseding
ties with friends, in-laws, or often, even spouses (Keefe
1988). As I have noted,
Hicks (1976) also finds some truth to the widely
noted "taciturnity" of mountain people, who perhaps
wisely develop nonverbal and nonemotional responses
to situations where others might be sensitive to insult,
and Appalachians may suffer discrimination by
cultivating these traits. For example, as a group they
tend to score higher in pathology on the MMPI, a
common psychological assessment tool, on the "social
introversion" scale that tests for "increasing levels of
social shyness, preference for solitary pursuits, and lack
of social assertiveness" (Keefe, Hastrup, & Thomas
1993:5).
Sometimes this results in interesting
visiting patterns within the nuclear
family. One woman I interviewed,
for example, married into her
husband's community. One of his
brothers lives nearby and the
brothers visit daily, often at her
house. However, she says, she
doesn't talk to her brother-in-law.
Her concern is directed more toward
her own "family," particularly her
parents with whom she visits every
Sunday. (Keefe 1988:26)
Cultural mechanisms for conflict resolution are
available in mountain churches where footwashing and
flower ceremonies are practiced.
Footwashing
commemorates the practice ordered by Christ when He
washed His disciples' feet prior to the Last Supper, and
involves all members of the congregation who move
forward one by one to wash another" s feet using a basin
of water and a towel. Howard Dorgan describes the
emotionally cathartic nature of this ceremony:
...one person finally decides to
approach, basin in hand, the one
church member for whom he or she
holds the most antipathy, or that
estranged family member toward
whom profound bitterness has in the
past been directed. These are the
encounters that typically produce the
highest drama and from which the
most needed therapeutics may be
derived. (Dorgan 1989:111)
Socializing is done mostly (sometimes exclusively)
with relatives. Holidays are always family-oriented,
and additional rituals such as family reunions and
Memorial Day gatherings in family cemeteries are
commonplace in Appalachia. Rural mountaineers may
be unfamiliar or ill-at-ease with forms of social
organization requiring interaction with strangers and
non-kin, such as the pervasive voluntary organizations
of mainstream America (Hicks 1976; Stephenson
1968).
Elsewhere, Dorgan (1987) describes a "Flower Service"
in which church members bring home-grown bouquets
and are urged to "Get right with thy neighbor" by
exchanging flowers symbolic of forgiveness of old
grudges and misunderstandings. He suggests that
lasting anger is difficult to maintain with this kind of
pressure for reconciliation. Yet, conflict is less easily
resolved beyond the boundaries of a particular church
or with newcomers unversed in mountain ways.
Moreover, there are few cultural mechanisms for
dealing with more overt hostility, and it can easily
escalate beyond the point that it can be handled by
normal mechanisms of social control.
Community in Appalachia is anchored on kinship
which provides the real social networks and the cultural
values that bind people into community (Beaver 1986).
And indeed households in the rural community are
connected by blood or marriage. Siblings continue to
maintain close ties into adulthood, and first cousins
also tend to be especially close. Matthews (1965), for
example, finds the naming of children follows
collateral ties to cousins and siblings rather than lineal
ties to parents or grandparents. Individuals may marry
cousins, including first cousins, which strengthens the
ties between collateral kin. Appalachian communities
often have sets of "double cousins," or children of
siblings who married siblings and, therefore, are related
to each other by double bonds of kinship (Matthews
1965).
Familism
The family is the fundamental social institution in
Appalachia, and familism is the basis for the
construction of social relations in general. The
Appalachian family is structured somewhat differently
Economy in Appalachia is a family economy which
is based on a relationship among kin who are
17
�associated with a specific plot of land (Beaver 1986;
Halperin 1990). Land is passed down through the
family, and it is land ownership that ultimately
connects family members (Bryant 1981). Family often
live on the same plot of land in Appalachia, and one
common manifestation of this is the clustering of small
homes and trailers housing family groups along
country roads. Family members rely on one another
for their well-being, exchanging goods and services
and cooperating in organized activities such as
gardening, farming, and house-building.
The
interdependence of kin networks and kin obligations is
manifest in the political sphere as well, where family
members' loyalty is expected in voting for agreed-upon
candidates, and where a candidate, if elected, is
expected to appoint kin to political offices (Hicks
1976).
throughout the Southern Appalachian ranges (Dorgan
1987). Churches often have "singings" during the
week when the congregation gathers for an hour or
more in the evening to sing hymns and play religious
tunes. During the summertime, singings may be held
outside and provide a social as well as sacred occasion.
A "Singing on the Mountain" is held each summer on
Grandfather Mountain, NC, attracting 10,000 or more
participants from near and far. A Trail of Tears
Singing is held in Cherokee, NC, each June
incorporating members of both Eastern and Western
Cherokee who, through Christian music, remember the
four thousand who died during the forced removal to
Oklahoma in 1838-39 (Neely 1991).
Mountain religion expresses the cultural themes
important to Appalachian people. Egalitarianism is
manifested in the simple church buildings housing
small, homogeneous congregations, led often by
preachers with only high school degrees who are
addressed not as "Pastor" but simply as "Brother soand-so."
Rituals such as creek baptism and
footwashing teach humility and simplicity. Concern
for individualism and personalism is found in the
emphasis by the mountain church on personal salvation
rather than on social works (Ford 1962). These values
are reinforced during church services at times when the
preacher singles out the personal sins of members to
illustrate the need to follow Biblical commandments, or
when the congregation joins in "oral prayer," each
individual saying their personal prayer aloud at the
same time as everyone else.
Appalachians continue to honor family members after
their death. This tradition contributes to an attitude of
greater acceptance of death in Appalachia than in
mainstream America, in which the denial of death is
more pronounced (Sherrod 1990). Attendence at both
the visitation preceding the funeral and the memorial
service is expected of family and friends (Crissman
1994). Funeral processions cause other drivers to pull
over spontaneously to the side of the road and stop
while they pass, a pattern that is followed only
sporadically for ambulances in my town! Deceased
relatives are frequently memorialized through the
newspaper publication of their pictures and special
poems written by living relatives (often years or
decades following the death), through Decoration
(Memorial) Day cleaning of cemetary graves and
placing of fresh flowers, and, as Dorgan notes, through
reverential readings of deceased members names
during church services.
The church is typically conceptualized as a
"family of Christ" (Bryant 1981). Fellow church
members are all "Brothers and Sisters" who worship
the Father. Cemeteries are maintained by those related
to the dead, as Appalachian poet Robert Morgan
reminds us in "Cleaning Off the Cemetery:"
Religious World View
Religion is pervasive in mountain life. Most
mountain people identify as Christians and practice
personal moral codes which are based on conservative
interpretations of the Bible, and which disavow
alcohol, gambling, and profanity. Even if they don't go
to church, mountaineers enjoy reading and arguing
about the Bible any day of the week (Humphrey 1984;
Miles 1975). Prayer is also a part of daily life. Until
recent Supreme Court rulings, prayers typically began
civic meetings and important rites of passage, such as
graduation. There is reliance on prayer for healing
sickness (Keefe & Parsons 1996). Rural Appalachians
favor country and gospel music, both of which
incorporate religious themes and references. Weekly
gospel programs are produced by small radio stations
Not the church-devout but those
reverent to family memory
show for these workings held
every three to five springs,
some driving a ways, complaining,
but always here on the chosen day
with tools and kids and dinner.
(Morgan 1987:52)
Familism is also evident in the congregations, which
are typically made up of members bound by ties of
social kinship, and in annual church "homecomings"
which celebrate the reunion of actual kin groups as well
as church members.
18
�James Still in the poem "Heritage" in which he declares
that "Being of these hills I cannot pass beyond" (Still
1968:82). Writers from the coalfields, including Still
(1940), Norman (1972), and Giardina (1987), typically
incorporate in their novels a sense of longing for the
land that has been stolen by the mining companies.
The landscape is also evoked spiritually as when poet
Robert Morgan states:
Sense of Place
Appalachian people have a strong sense that the
mountains are home. While outsiders appreciate the
southern highlands for aesthetic and economic reasons,
mountain people have sacred attachments to the land
that symbolizes family, livelihood, and ancestral roots.
A man in southwest Virginia captures this sentiment:
Land in this area doesn't change
hands much. The place we live on
has been in the family for seven
generations. This is not real estate,
this is home and will be handed
down, I hope, for seven more
generations. (Morefield 1990:63)
Mountains speak in tongues. Take
the wide thought of estuaries. The
absent god leaves the forest and the
tundra soaked in divinity. (Morgan
1978:42)
In mountain communities, this melding of the land,
the Lord, and the people is accomplished each year
when religious services are taken outdoors for
baptisms, Decoration Day, and tent revivals.
Mountaineers come to know their land not as a
generic but as a specific piece of earth with landmarks
and bits of local history that give it meaning (Hicks
1976). While newcomers feel compelled to leave the
mountains occasionally in order to cope with the
perceived isolation, mountain people love their land,
enjoy showing it off and being on it. When asked
where they would like go on a vacation, Wagner's
(1995) informants in southwest Virginia frequently
replied that they would like to take their vacation at
home!
Mountaineers feel the responsibility of stewardship
toward the land and the need to make it productive
while protecting and preserving it (Eller 1982). They
hope to inherit land from their parents, to work it, and
to pass it on to their children. Land becomes associated
with families, and the "homeplace" where ancestors
settled is kept in the family and is often the location of
the family reunions. The homeplace includes not only
the house but also the grounds and even the farmland
of past generations.
Land is essential to the
maintenance of the "multiple livelihoods" of mountain
families, and people struggle to keep their land, often
commuting 50-100 miles a day to find wage labor
(Halperin 1990). In this way, Appalachians are able to
maintain the rural lifestyle they prefer while taking
advantage of urban resources that are necessary.
Those who migrate out of the mountains commonly
complain of being homesick for the place as well as the
people, and there are high rates of return migration as
a result. A middle-aged woman from Mount Rogers,
VA, reminisces:
I moved away from here one time,
up north to Baltimore. Now I know
this is going to sound silly but —
when I would say my prayers at
night it was like "God" wasn't there.
I know that's crazy but that's the way
it felt to me. It was because I wasn't
in the mountains where I was
supposed to be — where I needed to
be. (Morefield 1990:23)
Mountaineers identify with their community which is
the integration of people and place. Typically, these
are small locales where residents know one another and
feel comfortable and secure. Mountain communities
are unlikely to be listed on any maps of the area, which
generally include only incorporated towns and county
seats. In one study, 72 communities were identified in
a single rural Appalachian county in North Carolina
(Plaut, Landis & Trevor 1993). These included rural
roads and hollars, family compounds, crossroads
hamlets, river and valley bottoms, and mountainside
clusters of homes. Residents have deep loyalty to their
place and a suspicion of other places. There is also a
general disdain for urban and industrial ways of life
(Halperin 1990).
Richard Humphrey finds this sense of the mountains as
a sacred place particularly strong in deep rural areas
where people are likely to practice what he calls the
"religion of Zion," Zion being the mountainous place in
the Old Testament referred to in Psalms 121:1-2: "I
will lift up mine eyes unto the hills from whence
cometh my help." Appalachian writers and poets often
refer to this kinship with the land felt by mountain
people, and perhaps none has captured it as well as
19
�Contemporary Ethnic Identity
"mountain people," "country people," or
"mountaineer." Regardless of the lack of agreement on
an ethnic name, Appalachian natives can usually
identify members of their group through the
recognition of an individual's dialect and his or her
claims to a homeplace and relatives in the mountains.
Appalachians are certainly able to tell you what group
they do not belong to, as illustrated in the variety of
pejorative names applied to non-Appalachians:
Yankee, outsider, "touron" (tourist/moron), "Floron"
(Florida resident/moron), and "Floridiot" (Florida
resident/idiot). The last three labels derive from
gawking and insensitive tourists and seasonal residents
who flood the southern mountains for relief from the
heat in the summertime. Through careful employment
of this oppositional identity, mountaineers tend to
continually negotiate their ethnicity, moving in and out
of an ethnic group status as they vie for resources and
autonomy. Foster (1988) demonstrates this process of
shifting ethnicity as native residents of Ashe County,
NC, vocally assume a distinctive identity as people
with an endangered "mountain culture" in order to stop
federal authorities from damming the New River and
destroying farmlands. At the same time, residents are
not completely comfortable with this identity because
of the inferior status conferred on ethnic minority
groups in America.
Conclusion
While Appalachian people are culturally similar to
many others in rural America, especially the rural
South, they retain a special identity connected to the
mountains. This identity comes in part from an
awareness of the differences between "us and them," an
awareness gained most quickly by Appalachians who
migrate out of the region and are forced to contend
with prejudice and discrimination in cities (Appalshop
Films 1983). But these differences are also becoming
apparent to Appalachian natives who stay in the
mountains, as more and more tourists and newcomers
arrive who feel little in common with mountaineers. I
moved to western North Carolina almost two decades
ago, and my uncle from Idaho came to visit after my
daughter was born. I felt sure he would feel
comfortable in the mountains since I noticed life was so
similar to the rural area in which I was born and raised
in the Northwest. But, instead, he was struck by how
different the people were in Appalachia, particularly
the way they talked, and he couldn't wait to get back
home.
The Appalachian dialect is one of the five or six
distinctive regional dialects recognized by linguists in
the United States (Wolfram and Fasold 1974). In one
of the few comprehensive studies of "Appalachian
English," Wolfram and Christian (1976) find distinctive
phonological and grammatical features, including
retaining the initial H in auxiliaries and pronouns ("hit"
for "it"), the use of double modal verbs ("I might could
do that"), and the use of variant pronouns ("hisself' for
"himself1 and "you'uns" for "all of you"). Social
linguists observe that dialect is a primary means by
which boundaries of social groups are marked,
especially among people otherwise culturally similar.
As one of my informants said, "People know who I am
the minute I open my mouth." The stigma against
"country English" in cities and mainstream America is
something that every Appalachian person must contend
with in the process of forming a personal identity. By
adopting a conscious pride in their cultural heritage,
Appalachians are better able to withstand negative
images held by others. Some are able actually to
exploit the differences, such as best-selling author, Lee
Smith, who declares: "The mountains which used to
imprison me have become my chosen stalking ground"
(Smith 1994).
Appalachian ethnicity has emerged out of more than
a century of "cultural reflexivity" during which the
people have been forced by colonizers and political
foes to question their cultural identity and their right to
preserve their cultural heritage (Roosens 1989). Their
ethnicity has gained strength in an era in which class
struggle has become increasingly less successful in
organizing in relation to multi-national industries and
federal agencies. As ethnics, Appalachians can take
advantage of legal precedents and global movements
claiming equality and equal treatment for ethnic groups
perceived to have the right to retain their culture and
define what it consists of (Roosens 1989).
As "reluctant ethnics," Appalachians will continue to
negotiate the tensions produced by the inherent
contradictions of the core American values of equality
(or sameness) and individualism (or differentness)
(Hicks & Handler 1988). This tension creates a
powerful force in American society especially when
these values are consistently at odds with the social
reality of ethnic inequality and the power of group
labelling. Given their fundamental similarities in race,
language, and national origin to white mainstream
Americans, mountain people have the option of
For Appalachians, there is little consistency in ethnic
label identification. The term "Appalachian" is one
used by scholars but only rarely assumed by members
of the group. More common are identities as
20
�Edward J. Cabbell. 1985. "Black Invisibility and Racism in
Appalachia: An Informal Survey," in Blacks in Appalachia,
eds. William H. Turner and Edward J. Cabbell. Lexington:
University Press of Kentucky.
John C. Campbell. 1921. The Southern Highlander & His
Homeland. New York: Russell Sage Foundation, p. 12.
adopting or resisting ethnic status as it suits their need.
As ethnics, they are able to organize, to manipulate
symbols of their common heritage, and often to be
successful in influencing political policy and the
distribution of economic goods. As non-ethnics, they
may distance themselves from ascribed ethnic traits,
"pass" into the mainstream of American society, and
achieve the American Dream in the "melting pot." This
strategy of reluctant ethnicity can best be understood as
a reasonable Appalachian response to the American
context in which individual and group differences are
at once denied and celebrated.
James K. Crissman. 1994. Death and Dying in Central
Appalachia: Changing Attitudes and Practices. Urbana, IL:
University of Illinois Press.
Howard Dorgan. 1987. Giving Glory to God in Appalachia.
Knoxville, TN: University of Tennessee Press.
Howard Dorgan. 1989. The Old Regular Baptists of Central
Appalachia: Brothers and Sisters in Hope,, Knoxville, TN:
University of Tennessee Press.
References
Ronald D. Eller. 1982.
Miners, Millhands, and
Mountaineers: Industrialization of the Appalachian South,
J880-1930. Knoxville, TN: University of Tennessee Press.
Richard D. Alba. 1990. Ethnic Identity: The Transformation
of White America. New Haven: Yale University Press, p. 342.
The Appalachian Regional Commission. 1965. Appalachia^
Washington, DC: U.S. Government Printing Office.
Judith Ivy Fiene. 1993. "The Appalachian Social Context
and the Battering of Women," Practicing Anthropology,
15(3) summer, 20-24.
The Appalachian Land Ownership Task Force. 1983. Who
Owns Appalachia?
Landownership and Its Impact,
Lexington, KY: University Press of Kentucky.
David Hackett Fischer. 1989. Albion's Seed: Four British
Folkways in America. New York: Oxford University Press.
Allen Batteau. 1990. The Invention of Appalachia. Tucson,
AZ: University of Arizona Press.
Thomas R. Ford, ed. 1962. The Southern Appalachian
Region: A Survey. Lexington, KY: University of Kentucky
Press.
Patricia D. Beaver. 1986. [reissued 1992]. Rural Community
in the Appalachian South. Prospect Heights, IL: Waveland
Press.
Thomas R. Ford. 1962. "The Passing of Provincialism," in
The Southern Appalachian Region: A Survey, ed. Thomas R.
Ford Lexington, KY: University of Kentucky Press.
Patricia D. Beaver and Darlene Wilson. March, 1997. "To
Embrace the Male Offshore Other: The Ubiquitous Native
Grandmother in America's Cultural History," presented at the
annual meeting of the Appalachian Studies Association, Fort
Mitchell, KY.
Stephen William Foster. 1988. The Past is Another Country:
Representation, Historical Consciousness, and Resistance in
the Blue Ridge. Berkeley, CA: University of California
Press.
Dwight Billings, Kathleen Blee, and Louis Swanson. winter
1986. "Culture, Family, and Community in Preindustrial
Apppalaehia," Appalachian Journal, 13, 154-170.
Denise Giardina. 1987. Storming Heaven: A NoveL New
York: Norton
Rhoda H. Halperin. 1990. The Livelihood of Kin: Making
Ends Meet "The Kentucky Way." Austin, TX: University of
Texas Press.
Ora Blackmun. 1977. Western North Carolina: Its
Mountains and Its People to 1880. Boone, NC: Appalachian
Consortium Press.
George L. Hicks. 1976 [reissued 1992]. Appalachian Valley.
Prospect Heights, IL: Waveland Press,
H. Tyler Blethen. 1994. "The Transmission of Scottish
Culture to the Southern Back Country," in Appalachian
Adaptations to a Changing World, Journal of the
Appalachian Studies Association, ed. Norma Myers, 6, 5972.
George L. Hicks and Mark J. Handler. 1987. "Ethnicity,
Public Policy, and Anthropologists," in Applied
Anthropology in America* second edition, eds. Elizabeth M.
Eddy and William L. Partridge New York: Columbia
University Press.
F. Carlene Bryant. 1981. We're All Kin. Knoxville, TN:
University of Tennessee Press.
21
�Dana Kovarsky, Toby Stephan, and Maria Braswell. 1994.
"Conflict Talk in an Appalachian Day Care Center," in
School Discourse Problems, second edition, eds. Danielle
Newberry Ripich and Nancy A. Creaghead San Diego:
Singular Publishing.
Helen Matthews Lewis, Linda Johnson, and Donald Askins.
1978. Colonialism in Modern America: The Appalachian
Case. Boone, NC: Appalachian Consortium Press.
Charles M. Hudson, ed. 1985. Ethnology of the Southeastern
Indians: A Sourcebook. New York: Garland Pub.
Richard Humphrey. 1984. presentation at Appalachian State
University, Boone, NC.
Richard Humphrey. 1984. "Religion and Place in Southern
Appalaehia," in Cultural Adaptations to Mountain
Environments, Southern Anthropological Society
Proceedings, No. 17, eds. Patricia D. Beaver and Burton L.
Purrington Athens, GA: University of Georgia Press.
Ronald L. Lewis. 1987. Black Coal Miners in America:
Race, Class, and Community Conflict, 1780-1980. Lexington,
KY: University Press of Kentucky.
John C. Inscoe. 1989. Mountain Masters: Slavery and the
Sectional Crisis in Western North Carolina. Knoxville, TN:
University of Tennessee Press.
ElmoraMesser Matthews. 1965. Neighbor and Kin: Life in
a Tennessee Ridge Community., Nashville, TN: Vanderbilt
University Press.
Loyal Jones. 1994. Appalachian Values. Ashland, KY: The
Jesse Stuart Foundation.
Gordon B. McKinney. 1978.
Southern Mountain
Republicans, 1865-1900; Politics and the Appalachian
Community. Chapel Hill, NC: University of North Carolina
Press.
Susan E. Keefe. 1985. unpublished field notes.
1994. "Urbanism Reconsidered: A Southern
Appalachian Perspective," City and Society,
Annual Review, 1, 20-34.
Emma Bell Miles. 1975. The Spirit of the Mountains. New
York: J. Pott, 1905; reprinted Knoxville, TN: University of
Tennessee Press.
1992. "Reluctant Ethnics: The Interplay of
Equality and Individualism in Southern
Appalachia," paper presented at the annual meeting
of the American Anthropological Association, San
Francisco.
Teena Morefield. 1990.
"The Appalachian Culture:
Implications for Counselors," unpublished paper, p. 63.
Robert Morgan. 1987. "Cleaning Off the Cemetary," in At
The Edge of the Orchard Country. Middletown, CT:
Wesleyan University Press, p. 52.
1988. "Appalachian Family Ties," in Appalachian
Mental Health, ed. Susan Emley Keefe Lexington,
KY: University Press of Kentucky.
Robert Morgan. 1978. "Mockingbird," in Trunk and Thicket.
Fort Collins, CO: L'Epervier Press, p. 42.
Susan E. Keefe and Paul Parsons. March 30, 1996. "A
Survey of Appalachian and Non-Appalachian Health and
Lifestyles Indicators in Watauga County, NC," presented at
the annual meeting of the Appalachian Studies Association,
Unicoi, GA.
Sharlotte Neely. 1991. Snowbird Cherokees: People of
Persistence. Athens, GA: University of Georgia Press.
Gurney Norman. 1972. Divine Right's Trip, New York: Dial
Press
Susan E. Keefe, Mark D. Vickery, and Karen Dunlap. 1984.
"Watauga County Churches," unpublished ms.
Marion Pearsall. 1966.
"Communicating with the
Educationally Deprived," Mountain Life cfe Work, 42(1), 811.
Susan E. Keefe, Janice L. Hastrup, and Sherry F. Thomas.
November 21, 1993. "Psychological Testing in Rural
Appalachia," presented at the annual meeting of the
American Anthropological Association, Washington, DC.
William Petersen. 1980. "Concepts of Ethnicity," in Harvard
Encyclopedia of American Ethnic Groups. Cambridge, MA:
Harvard University Press.
Susan E. Keefe, Gregory G. Reck, and Una Mae Lange Reck.
1989. "Measuring Ethnicity and Its Political Consequences
in a Southern Appalachian High School," in Negotiating
Ethnicity: The Impact of Anthropological Theory and
Practice, ed. Susan Emley Keefe. Washington, DC:
American Anthropological Association, National Association
for the Practice of Anthropology, Bulletin No. 8.
Thomas Plaut, Suzanne Landis, and June Trevor. 1993.
"Focus Groups and Community Mobilization," in Successful
Focus Groups: Advancing the State of the Art, ed. David L.
Morgan Newbury Park, CA: Sage Press.
Mary Beth Pudup. winter 1990. "The Limits of Subsistence:
Agriculture and Industry in Central Appalachia," Agricultural
History. 64. 61-89.
Susan E. Keefe, Una Mae Lange Reck, and Gregory G. Reck.
1983. "Ethnicity and Education in Southern Appalachia: A
Review," Ethnic Groups, 5, 199-226.
22
�US Bureau of the Census. 1990 Census of Population,
Supplementary Reports, Detailed Ancestry Groups for States.
Karl B. Raitz and Richard Ulack. 1991. "Regional
Definitions," in Apyalachia: Social Context Past and
Present, Third edition, eds. Bruce Ergood and Bruce E.
Kuhre. Dubuque, IO: Kendall/Hunt.
Melinda Bollar Wagner, Shannon T. Scott, Megan Scanlon,
Stacy L. Viers, and Jean A. Kappes. 1995. "It May Not Be
Heaven, But It's Close:" Land and People in Craig County,
Virginia," Radford, VA: Radford University Appalachian
Regional Study Center.
Eugeen E. Roosens. 1989. Creating Ethnicity: The Process
ofEthnogenesis. Newbury Park, CA: Sage,
Anya Peterson Royce. 1982. Ethnic Identity: Strategies of
Diversity. Bloomington, IN: Indiana University Press.
Altina L. Waller. 1988. Feud: Hatfields, McCoys, and
Social Change in Appalachia, 1860-1900. Chapel Hill, NC:
University of North Carolina Press.
Henry D. Shapiro. 1978. Appalachia on Our Mind: The
Southern Mountains and Mountaineers in the American
Consciousness, 1870-1920. Chapel Hill, NC: University of
North Carolina Press, p. x.
Mark Sherrod. 1990. "Asleep in Jesus: Death Rituals in
Southern Appalachia,"
unpublished M.A. thesis,
Appalachian State University.
David Walls. 1978. "Internal Colony or Internal Periphery?
A Critique of Current Models and an Alternative
Formulation," in Colonialism in Modern America: The
Appalachian Case, eds. Helen Matthews Lewis, Linda
Johnson, and Donald Askins Boone, NC: Appalachian
Consortium Press.
Herb E. Smith. 1983 Strangers & Kin. Whitesburg, KY:
Appalshop Films.
Mary Waters. 1990. Ethnic Options. Berkeley: University of
California Press.
Lee Smith. September 8,1994. "The Terrain of the Heart,"
convocation address at Appalachian State University, Boone,
NC, p. 12.
Jack E. Weller. 1965.
Yesterday's People:
Contemporary Appalachia. Lexington, KY:
University of Kentucky Press.
JohnB. Stephenson. 1968. Shiloh: A Mountain Community.
Lexington, KY: University of Kentucky Press.
David E. Whisnant. 1983. All That is Native & Fine. Chapel
Hill, NC: University of North Carolina Press.
John B. Stephenson and L. Sue Greer. 1983. "Ethnographers
in Their Own Cultures: Two Appalachian Cases," Human
Organization, 40, 123-130.
Eliot Wigginton and Margie Bennett, eds. 1984. Foxfire 8.
New York: 1st Anchor Books.
Life
in
Charles Williams. 1982. "The Conversion Ritual in a Rural
Black Baptist Church," in Holding on to the Land and the
Lord: Kinship, Ritual, Land Tenure, and Social Policy in the
Rural South, Southern Anthropological Society Proceedings,
No. 15, eds. Robert L. Hall and Carol B. Stack Athens, GA:
University of Georgia Press.
James Still. 1986. "Heritage," in The Wolfpen Poems.
Frankfort, KY: Berea College Press, distributed by Gnomon
Distribution, p. 82.
1940 River of Earth. New York: The Viking
Press.
J.W.Williamson. 1995. Hillbillyland: What the Movies Did
to the Mountains and What the Mountains Did to the Movies.
Chapel Hill, NC: University of North Carolina Press.
Ivan Tribe. 1991. "Traditional Appalachian Music/Early
Commercial Country Music: Continuity and Transition," in
Appalachia: Social Context Past and Present, Third edition,
eds. Bruce Ergood and Bruce E. Kuhre. Dubuque, IO:
Kendall/Hunt.
Margaret R. Wolfe. 1980-81. "The Appalachian Reality:
Ethnic and Class Diversity," East Tennessee Historical
Society's Publications, 52, 40-60.
William H. Turner. 1985. "The Demography of Black
Appalachia: Past and Present," in Blacks in Appalachia, eds.
William H. Turner and Edward J. Cabbell, Lexington, KY:
University Press of Kentucky, p. 257.
Walt Wolfram and Donna Christian. 1976. Appalachian
Speech, Arlington, VA: Center for Applied Linguistics.
William H. Turner. 1985. "Introduction," in Blacks in
Appalachia, eds. William H. Turner and Edward J. Cabbell,
Lexington, KY: University Press of Kentucky, p. xix.
US Bureau of the Census. 1992. 1990 Census of Population,
Supplementary Reports, Detailed Ancestry Groups for States,
Washington, DC: US Government Printing Office.
23
�Appalachia as defined by the
Appalachian Regional Commission in 1996:
399 counties in 13 states
�II. Exploring a Region through Quantitative Data
Frequencies and Percentages
Welcome to Appalachia and the world of social research. You have new technology at your
fingertips to explore the region's geography, settlement patterns, successes and problems. You will
be using a "student version" of the MicroCase statistics package with real data gathered from a
variety of government and private agency sources.
The MicroCase software enables you to conduct professional analyses and yet it is easy to use—
you'll master its basic commands in a few minutes. By the end of this workbook, you should have
a good understanding of the Appalachian Region, be able to conduct your own research, have a
solid introduction to the uses of quantitative statistics and be comfortable using computers for
research and data analysis.
GETTING STARTED
The MicroCase statistics software requires an IBM-PC or fully compatible computer with 640K of
memory (RAM) and a VGA graphics card.
To use the MicroCase program the first time, it's helpful to pair up with a friend or fellow student.
One of you can operate the computer while the other reads the instructions. After the first few
minutes, you'll be comfortable enough with the program to work individually.
To begin, place the disk in the 3.5 inch floppy drive—the A or B drive. Your computer monitor
should be displaying the "C prompt," which looks like "C:>". Type A: (or B: if the disk is in the
B drive) and press <ENTER>. The computer will change the prompt from the C drive to the A or
B drive. When you see A: (or B:) type MC and press <ENTER>. (If you run the program from a
Windows environment, you may need to type MCIII and press <ENTER>.)
Once you have typed MC and pressed <ENTER>, it will take about 30 seconds for the program to
load. The first time you start Student MicroCase, you will be asked to enter your name. It is
important to type your name correctly, since it will appear on all printed output. Type in your name
and press <ENTER>. If it is correct, simply press <ENTER> in response to the next prompt. (If
you need to make a correction, type Y at the prompt and press <ENTER>. The copyright screen
will appear. Press <ENTER> to continue to the first of two primary menus. Its border will be blue
and labeled DATA AND FILE MANAGEMENT. The menu shows all the choices available with
the complete MicroCase statistics package (A through P). The choices you are allowed to make
with this abbreviated "student version" of the MicroCase Analysis System are in a darker print and
starred, for example
*L Open, Look, Erase or Copy File. This is the place to start and you will see that it is
automatically marked with a box. (You can move the box with the arrow keys.)
25
�The menu looks like this:
DATA AND FILE MANAGEMENT
*S. Switch to STATISTICAL ANALYSIS MENU
DATA MANAGEMENT
A. Define Variables/Recedes
B. Collapse/Strip Categories
C. Enter Data from Keyboard
D. List or Print Variable Values
*E. Codebook
F. Edit Variable Information
G. Grading Recode
H. Setup Data Entry
FILE MANAGEMENT
*!. Open, Look, Erase or Copy File
J. Create New Data File
K. Create Subset File
L. Import/Export Data
M. Move Data between Files
N. Merge Files
O. Create Aggregation File
P. Create Statistical Summary
*X. EXIT from MicroCase
OPEN FILE:C:\APCOUNTY
Note that I. Open, Look, Erase or Copy File is highlighted by a box around it. Press <ENTER>
and you will see a single file: APCOUNTY. Press <ENTER> again and you will see a new screen
with "File Name:C:\APCOUNTY" at the top. Below is a description of the file: "Data on the 399
Federally-designated Appalachian Counties" and a list including the Current Number of Cases: 399.
Each of the federally designated Appalachian Region's 399 counties is a case. (The county is the
unit of analysis in this data base; other studies might use census tracts, states or nations as the unit
of analysis. Surveys use individuals as the unit of analysis.) Next you will see the number of
variables: 90. You will be able to compare each of the 399 counties across these 90 variables. For
example, you can compare counties in terms of how much money people make (per capita income),
average levels of education, percentage of the labor force engaged in mining, percentage of voters
who favored Clinton in 1990, etc. Press <ENTER> to continue. Note that the functions that can be
used with this trimmed-down version of the software are starred (*), You need the full MicroCase
Analysis System to access all the functions. Press the S. Switch to STATISTICAL ANALYSIS
menu function (which should have the highlighting box around it) to obtain a listing of statistical
procedures. The STATISTICAL ANALYSIS menu has a red border to clearly differentiate it from
the blue-bordered DATA FILE AND MANAGEMENT menu. You move between the two menus
by pressing S. The STATISTICAL ANALYSIS menu looks like this:
26
�STATISTICAL ANALYSIS
*S. Switch to DATA AND FILE MANAGEMENT MENU
DATA MANAGEMENT
*A. Univariate Statistics
*B. Tabular Statistics
*C. Analysis of Variance
D. Covariance Analysis
*E. Mapping Variables
*F. Scatterplot
*G. Correlation
H. Partial Correlation
*l. Regression
ADVANCED STATISTICAL ANALYSIS
J. Regession Models
K. Curve Fitting
L. Factor Analysis
M. Logistic Regression
N. Time Series
Q. Interactive Batch
*X. EXIT from MicroCase
OPEN FILE:C:\APCOUNTY
OK. You're ready to start exploring Appalachia. Press E. Mapping Variables. At the prompt
"Enter name or number of variable to be mapped:" type 2 and press <ENTER>. A map of the
Appalachian region will unfold across the screen, state by state. Colors will be added and you will
be able to see Appalachia's 399 counties spread across portions of 12 states and all of West
Virginia. The Federal Government has determined that Appalachia consists of 399 counties in 13
states. Press D (for distribution) and you will see a list of the counties and states. (Don't worry
about the numbers to the right of the county names-I used them here only to color code the states
so they would stand out from each other on the map.) Press the Page Down key to scroll down the
399 counties and make a list of Appalachia's thirteen states;
1
8 _
2
9 _
3.
10.
4,
11
5_
12.
6
13
27
�Press <ENTER> to return to the map and press <ENTER> again to get the variable prompt. Type
3 and press <ENTER>, You will see Appalachia divided into three regions: Northern, Central and
Southern. In 1974, Appalachian Regional Commission (ARC) planners decided these regions have
different strengths and needs that required different strategies for development:
The Northern region (coded yellow) includes portions of New York, Pennsylvania,
Ohio, Maryland and all but the nine most southeastern counties of West Virginia.
Seen as an area having an old or outmoded manufacturing economic base, the 1990
per capita income was $17,265 or 83.6% of the U.S. average, $20,652.
The Central Appalachian region (orange) contains the bituminous coal fields and
timber. The ARC said that the region had been "plagued for decades" with socioeconomic deficiencies. 1990 per capita income was only $13,073 (76% of the
Northern Appalachian subregion and 63.3 % of the U.S. average). Why should an
area so rich in resources be so poor? The commission cited the rough mountain
terrain and an economy focused on the coal industry (Appalachia magazine, Vol.
8, No. 1, August-September 1974, pp. 10-27). Some critics, such as Jack Weller in
his book Yesterday rs People, saw problems in the values and world views of
mountain people, while others, such as Helen Lewis (see "Fatalism or the Coal
Industry" in Ergood and Kuhre's Appalachia; Social context, Past, Present and
Future) traced the problem to damage done by extractive (coal and timber)
industries and corporations and government policy. Were the region's problems
caused by the values of its residents, or by historical events, such as
industrialization? The data you will explore in the coming chapters should help you
formulate your own opinion.
The Southern region (red), running from the Blue Ridge of Virginia to Mississippi,
was described by ARC planners as an area "moving from an agrarian-based
economy to a new, modern industrial economy with rapid population growth being
the outstanding characteristic of the region." The ARC called for balanced growth
between urban and rural areas, diversification of employment opportunities, and a
new leadership for economic development. Its 1990 per capita income was $ 16,893
or 81.8% of the U.S. average.
Let's see how many counties are in each of the three regions. Press <ENTER> until you return to
the STATISTICAL ANALYSIS menu. Select A. Univariate Statistics. At the "Enter the name
or number of the variable" prompt, type 3 and press <ENTER>. Bypass the next request for a subset
variable by pressing <ENTER> and you will see a "pie chart" containing three slices, each
representing one of the three regions. Press the up or down arrow key and you will see a slice
"exploded" away from the rest of the pie and identified as "North 144 36.1%," "Central 84 21.1%"
or "Southern 171 42.9%." Translated, this means the Northern subregion as having 144 counties,
or 36.1 percent of the total of Appalachia's 399 counties. The Central subregion has 84 counties
or 21 percent and the Southern has 171 counties or 42.9 percent. Press T and you will see a table.
Fill in the blanks (ignoring the columns listing cumulative percent (Cum.%) and Z-Score):
28
�Table 1: Number of Counties in the Appalachian Sub-Regions
%
Frequency
North
Central
Southern
Tables such as this one usually add another row for totals and look like this:
Table la; Number of Counties in the Appalachian Sub-Regions
Frequency
%
North
Central
Southern
Total
399
100
Press <ENTER> twice to return to the STATISTICAL ANALYSIS menu and select E. Mapping
Variables. Press the F3 function key at the top of your keyboard. A list of all 90 variables in the
APCOUNTY data file will appear. You can use the arrow keys to scroll down the list, the page
down/up keys to jump down a full screen of 17 variables and the home and end keys to move to the
beginning and the end of the list. You can either enter a variable's number at the prompt (as you
have already done in previous exercises), or you can use the left arrow key to mark a variable.
Practice this by pressing F3 to obtain the list of 90 variables. Arrow down to 5) Highlands on the
F3 variable list and press the left arrow to mark it. Press <ENTER> to see a map of those counties
designated as "highlands" by ARC planners, who defined them as areas over 1,000 feet above sea
level, which consequently have good potential for recreational development. Press F3 again to see
that you can obtain the variables list as an overlay on the map.
Okay, let's see if you can repeat the process you did for the three Appalachian subregions and
create a table showing highland and non-highland counties. Press <ENTER> three times to return
to the STATISTICAL ANALYSIS menu. Press A. Univariate Statistics. At the variable prompt,
press F3, arrow down to 5)Highlands, press the left arrow to mark it and press <ENTER> twice
to bypass the request for a subset. (Don't worry about subsets now; they will be explained in
chapter 5.) The pie chart should look like an environmentally (politically?) correct scene, portraying
a blue sky above a slightly tilted green mountain. Matching the color code legend in the upper left
of the monitor screen with the pie chart, you can see that a majority of Appalachian counties are
not highland counties. Press T (for Table) and fill in the following blanks:
29
�Table 2. Highland and Non-Highland Appalachian Counties
Frequency
%
Not High
Highlands
Total
Well done! In these first few pages you have learned how to define Appalachia and its three
subregions. You've learned how to access the MicroCase software package, open a data file, make
maps and simple tables to describe univariate (one variable) data. In future chapters, you will learn
how to compare two variables (bivariate data) or several variables (multivariate data). Simple
practice will enable you to move about the MicroCase software comfortably. Practice—playing with
data-is central to learning the basic elements of data analysis and how to use the software. The
more you play, the more you 77 know. It's similar to learning how to ride a bike or drive a car.
At this point I need to explain that statistical maps are not usually used to block out subregions of
an area. It's been a good way to introduce you to the software and to Appalachia, but the mapping
function is really used to portray the distribution of variables: For example, where is unemployment
the highest (and the lowest)? Press <ENTER> twice to return to the STATISTICAL ANALYSIS
menu and select E. Mapping Variables. Press F3 and type S. At the Phrase prompt, type
"unemployment" and press <ENTER>. A second box appears, listing two variables related to
unemployment: 37)UNEMP91 and 38)UNEMP80. Use the left arrow to mark 37)UNEMP91 and
press <ENTER> twice to obtain a map of 1991 CIVILIAN LABOR FORCE UNEMPLOYMENT.
Take a tour of the black prompts bar at the bottom of the screen. Dist. stands for distribution of
cases, from the highest rate or percentage to the lowest. (You used this function previously to
discover the states with counties in Appalachia, but, technically, it was not a correct use of the
function because there was no rank ordering of states from high to low—it was just a simple list.)
If you press D, you will find that Graham County, North Carolina has the highest unemployment
rate (25.7%), followed by Elliott County, Kentucky (22.6%). The screen displays the highest 50
cases. Since there are 399 cases over all, you would have to press the Page Down function key
seven times to find the county with the lowest unemployment rate, which is Tompkins County, NY
(3.4 percent). Press A (for area) to return to the map andpress L for the map's legend, showing the
color-coded breakdown of the distribution into five clusters, ranked from lowest to highest scores.
Press <ENTER> to return to the map andpress S for spot, which recreates the map with spots and
colors depicting unemployment rates. Messy, isn't it? With so many cases (399), spot maps aren't
very helpful. Comp allows the comparison of maps for two variables; we will do this in the next
chapter. Name points to the county with the highest rate; you can move down the list by using the
down arrow key. "Print" enables the printing of maps.
30
�For practice, use the F3 key and the mapping and distribution functions to discover which
of the three regions is the most rural
of the three regions has the most mining
county has the highest mean per capita income
county has the highest level of education
county has most doctors per person
Which Appalachian county is the most interesting to you? Why?
Welcome to Appalachia, and a computerized statistical package that enables exploration of a region
similar in many ways to the nation that surrounds it, while also maintaining its own unique
characteristics.
Throughout this book, you will be asked to translate data into your own words. Describing data is
a skill that requires practice to learn. Begin this learning process now by writing summary
statements about
Appalachia and its constituent states and number of counties
Appalachia's three regions and their differences
31
�The computer maps and what they can demonstrate.
What frequencies can tell you.
Why percentages are more helpful in explaining data than frequencies.
Some students enjoy summarizing data with computer slide presentation software, such as
Microsoft's PowerPoint or Corel's Presentations. You may want to experiment with one of these
programs as a means of presenting and explaining data.
32
�Appalachian Settlement Patterns:
Discovering a Multi-Cultural Region
1990 Census: Percent claiming Scotch-Irish Ancestry
�III. Cultural Diversity in Appalachia
Mapping, Rank Ordering and Correlation Coefficients
Appalachia enjoys a rich multi-cultural diversity. There are Native Americans and African Americans,
Scotch-Irish, German and Italians. Ever heard of the Pennsylvania Dutch? They're not Dutch. They're
Deutsche—German. Let's see where the Germans settled in Appalachia.
Open the APCOUNTY file, switch to the red STATISTICAL ANALYSIS MENU. Place the highlight on
E. Mapping Variables and press <Enter>. Now the screen asks you for the variable you want to map. Type
13 or %German90. What you are mapping is the percent of people in a county who reported any German
ancestry in the 1990 census. Type N (for Name) to see which county has the highest percentage of people
claiming German ancestry. Snyder, Pa. has the highest percentage: 66.07%. Press the down arrow to find
the county with the next highest rate. It is Holmes, Ohio, with a rate of 64.86%. Press <ENTER> to return
to the map mdpress D (for distribution). You will see a list of all 399 counties, listed from the county with
the highest percent to the lowest. Snyder, Pa. should be at the top. List the top five counties:
1. Snvder.Pa.
2.
3.
4.
5.
Press the DOWN ARROW repeatedly (seven times) to scroll down to the bottom of the list of 399 counties.
You will find that Benton County, Mississippi has the fewest people claiming German ancestry. List the
percentage of people claiming German ancestry in Benton County:
.
Notice that there is a color coding of the county names. The entire list is broken into fifths, each with its own
color. Press <ENTER> to return to the map and type L (for Legend). The legend indicates the ranges used
for clustering the counties into five groups (quintiles) of 80 counties each (except for the lowest group which
has 79—with 399 counties, one of the clusters had to get short changed.) The highest quintile has counties
with percentages of persons claiming German ancestry ranging from 32.90 to 66.87 percent. The lowest
quintile has percentages of persons claiming German ancestry ranging from 3.36 to 10.28 percent. Press
<ENTER> and look at the map again. Notice how the color coding provides you a quick, graphic view of
where Germans settled in Appalachia.
The Scotch-Irish also settled in Appalachia. Return to the red STATISTICAL ANALYSIS MENU. Place
the highlight on E. Mapping Variables and press <Enter>. Type 15 or %SCT.IRH90. Do you see any
clusters of Scotch-Irish? Press D (for distribution) and list the ten counties where the highest percentage of
people claim Scotch-Irish ancestry:
34
�L
2,
1
4.
5,
6,
1
JL
10,
_
In which state did the most people claim Scotch-Irish heritage?
Who are the Scotch-Irish? What can you learn about their origins and the routes they traveled into the
mountains? One good source is John C. Campbell's chapter on "Pioneer Routes of Travel and Early
Settlements" in his classic The Southern Highlander and His Homeland. (His chapter is reprinted in the
Ergood and Kurhre reader cited in the Resources listing on page 113).
35
�What does Susan Emley Keefe say in Chapter 1 about the ethnic identity of the Scotch-Irish, and why
might there may be a significant undercount in the numbers of Scotch-Irish claiming their heritage in the
1990 census?
The first inhabitants of Appalachia were Native Americans. They arrived in Appalachia long before the
Germans, Scotch-Irish and other Europeans and they still thrive in the mountains today. Can you find the
counties and states where descendants of Native Americans are clustered? Return to the STATISTICAL
ANALYSIS menu. Place the highlight on E. Mapping Variables and press <Enter>. Type 11 or
%AMER.IN90. Do you see any clusters of Native Americans? Press D (for distribution) and list the five
counties where the highest percentage of people claim such ancestry:
1.
2.
3.
4.
5.
What state is most represented?
Does this state rank high in terms of people claiming German ancestry?
Yes
No
Does this state rank high in terms of people claiming Scotch-Irish ancestry?
Yes
No
36
�Do you know what tribe is represented in the high ranked counties? It's the Cherokee. Any web browser such
as http://www.google.com/ will take you to web sites containing information on Cherokee history and
culture. Record below what you can find out about famous Cherokee leaders and important dates in the
history of the Cherokee people. One good web site is http://www.tolatsga.org/Cherokeel.html and
http://www.tolatsga.org/Cherokee2.html (This site also has a Cherokee2.html.)
Ever hear of the Trail of Tears? Read an eyewitness account by a soldier on the march at
http://www.powersource.com/cherokee/burnett.html. What did you learn from John G. Burnett's account
of the removal of the Cherokees from Western North Carolina?
37
�Few people realize that African Americans have a proud history in Appalachia. They built the railroads and
many worked in the coal mines. Return to the STATISTICAL ANALYSIS menu. Place the highlight on
E. Mapping Variables and press <Enter>. Type 18 or %BLACK90. Where is the black population? D (for
distribution) and list the five counties where the highest percentage of people claim African American
ancestry:
1.
2.
3.
4.
5.
What state is most represented?
Are Blacks and Germans clustered in the same parts of Appalachia?
Yes
No
Clearly the states having the highest African-American population are in the Deep South, where the
plantation economy was driven by the engine of slave and tenant farmer labor. But Blacks also joined the
migration north to the coal fields. Type N (for Name) and use the down arrow to find the counties in West
Virginia, Pennsylvania and Ohio ranked with the highest percentage of Black population. List them:
L
_
2,
_
4.
5,
6,
38
�Did people of different nationalities cluster together in Appalachia? Return to the STATISTICAL
ANALYSIS menu. Place the highlight on E. Mapping Variables and press <Enter>. Now press F3 and
you will see a list of all the variables in the APCOUNTY data file. Note that variables 11-18 all deal with
the percentage of people claiming different ethnic or national ancestry. Type 12 or %ENGLISH90 and
press <ENTER> to find what portion of Appalachia has the most people claiming English ancestry. Look
at the map carefully and note where the English settled. Press C (for comparison) and, when you are asked
for the Name or number of variable for comparison, type 16 or %SCOTTSH90.
Do the two maps look the same?
Yes
No
Does it appear that the English and the Scottish settled in the same area?
Yes
No
In the top right-hand corner of the screen is the equation r =
"rff
. (Enter the number.)
for the
It is a
and
as
and
the
or
the
to
in
the
the
association:
a
and 0.70
in the 0.20-0.50
are
are
Put
1 (+1
the
to
a
and
no
way,
r
is a
r: for
and
are
and 1
to Of (for 1
If r is
of
the
39
1.
is
a
are
is no
�Press <ENTER> to clear the lower map and, at the prompt for a new comparison variable, type 15 or
%SCT.IRH90.
Do the two maps look the same?
Yes
No
Does it appear that the English and the Scotch-Irish settled in the same area?
Yes
No
r=
Press <ENTER> to clear the lower map and, at the prompt for a new comparison variable, type 11 or
%AMERJND90.
Do the two maps look the same?
Yes
No
Does it appear that the English and Native Americans settled in the same area?
Yes
No
r=
Use ff F3 ff function key to find the variables needed to determine the correlations between people claiming
English and Scottish ancestry
0.63
English and Scotch-Irish ancestry
English and Native American ancestry
German and Italian ancestry
German and Black ancestry
German and Native American ancestry
Would you say that...
English and Scottish people seem to have settled in the same areas?
Yes
Not really
Clearly not
(circle one)
There is a strong moderate - weak - nonexistent (circle one) correlation
between these two variables.
40
�English and Scotch-Irish people seem to have settled in the same areas?
Yes
Not really
Clearly not
There is a strong moderate - weak - nonexistent (circle one) correlation
between these two variables.
English and Native American people seem to have settled in the same areas?
Yes
Not really
Clearly not
There is a strong moderate - weak - nonexistent (circle one) correlation
between these two variables
German and Italian people seem to have settled in the same areas?
Yes
Not really
Clearly not (circle one)
There is a strong moderate - weak - nonexistent (circle one) correlation
between these two variables.
German and Black Americans seem to have settled in the same areas?
Yes
Not really
Clearly not
There is a strong moderate - weak - nonexistent (circle one) correlation
between these two variables.
German and Native American people seem to have settled in the same areas?
Yes
Not really
Clearly not
There is a strong moderate - weak - nonexistent (circle one) correlation
between these two variables.
41
�Now use the codebook function to determine the mean (average) percentage of people in various states
claiming different kinds of ethnic heritage. Press <ENTER> until you return to the STATISTICAL
ANALYSIS menu. Press S to switch to the blue DATA AND FILE MANAGEMENT menu. Select E.
Codebook and follow these instructions carefully:
When asked to "Select output device," type 1 and press <ENTER> (to send the data to the
screen). Type Y, when asked if you want to stratify (which in this case means break the data
down by the thirteen states having counties in Appalachia). Enter 89JSTATESCODE as the
stratifying variable. Press <ENTER> to accept 89)STATESCODE as the independent
variable. Press <ENTER> again to bypass the request for a subset.
When asked for a "list of variables to be included," type 12,13,15,18 and press <ENTER>.
You should see the first of four tables, which list "N" counties by each of the 13 states, together with their
mean (average) percentage of people claiming English ancestry. The state claiming the highest English
ancestry is
.
Take a moment to list the number of Appalachian counties in each state:
New York
Pennsylvania
Maryland
Ohio
West Virginia
Virginia
Kentucky
Tennessee
North Carolina
South Carolina
Georgia
Alabama
Mississippi
Total Appalachian Counties
Press <ENTER> to find the two states with the highest mean claims to German ancestry.
Those states are
and
Press <ENTER> to find the two states with the highest average citations of Scotch-Irish
ancestry. They are
and
.
Press <ENTER> to find the single state where people claim the highest AfricanAmerican ancestry, which is
.
42
�To summarize this chapter, you have learned
that Appalachia is not a homogeneous stronghold of Anglo-Saxon stock, but
rather a heterogeneous, multi-cultural region.
that different groups settled in different parts of the region: Germans in
Pennsylvania and Maryland, the Scotch-Irish in the Carolinas, etc.
that some ethnic groups tended to settle near each other.
how to make and read MicroCase maps and how to compare maps.
how to find distributions of cases.
how to interpret the correlation coefficient "r" as a measure of both the strength
and direction of an association between two variables.
how to use the codebook function to review data and break data down by units
such as states (or regions, etc.).
Make a presentation describing settlement patterns and ethnic diversity in the Appalachian region. If
available, use computer software such as Microsoft's PowerPoint or Corel's Presentations. If such programs
are not available, you can use newsprint and markers. As previously noted, John C. Campbell's chapter on
"Pioneer Routes of Travel and Early Settlements" in The Southern Highlander and His Homeland is an
excellent resource, as is Susan Emley Keefe's introduction to the region in Chapter 1.
43
�This page intentionally left blank
�IV. Demographics: Urbanization and Migration in Appalachia
Mapping and Regression
Demographics is the study of population and population movements. It looks for characteristics of
size, density and migrations of human populations. Now that you know something of the ethnic
diversity of the region, look for the concentrations of populations. Open the APCOUNTY file and
select A. Univariate statistics. At the prompt, type 6 and press <ENTER> twice, bypassing the
request for a subset. You will see a "pie chart" indicating the ratio of metro to nonmetro counties.
About a quarter of the pie is metro—25 percent. Press the down arrow key to "explode" the pie,
emphasizing the metro piece. This is a simple way of making a chart a little more dramatic for
presentation. Press Fto obtain a table which indicates that 75.4 percent of the 399 Appalachian
counties (N=301) are nonmetro. Some 24.6 counties (N=98) are metro.
as
are
), or
are
10
(1,2,3,4,5....100) and
can
in a pie
you
a pie
or bar
for
but not for
or
use
will
and
as
To get
this
1; 21-39=2,
the
7
7
up
85
2is
into five
. Mapping Variables. At the prompt, type 6 mdpress <ENTER>. You will see a two color
map indicating counties in (or not in) metropolitan areas variously called "MS As" for Metropolitan
Statistical Areas or variations on the same name, such as CMSA=Consolidated Metropolitan
Statistical Area. Definitions vary from census to census, but, in general terms, a metropolitan area
consists of a central city area of at least 50,000 people, together with any surrounding counties tied
to the city by economic and commuting patterns for work and services. If you compare the map you
see on your monitor to an atlas, you will find the clusters of counties that make up the metropolitan
areas around Birmingham, Ala; Atlanta, Ga; Greenville, SC; Roanoke, VA.; the Knoxville-Bristol
corridor in Tennessee; Charleston, W.Va., and Pittsburgh (and much of industrial Pennsylvania).
Metropolitan areas appear throughout the mountain region.
Press Comp. The screen will divide in two and take about 20 seconds to make a smaller map of
the Appalachian region. You will be asked for the "Name or number of variable for comparison:"
Enter 7. Another map will appear showing the counties by percentage of population living in urban
areas (places of 2,500 people or more). The counties with the highest urban population should be
among the counties listed as "MSA, etc." counties in the top map. Let's see if this works; Press
<ENTER> three times to return to the variable prompt and type 7. After the map appears press D
and write down the 10 counties with the highest percentage of population living in urban areas:
45
�Allegheny. PA. 95.9%
Press <ENTER> three times to return to the variable prompt and type 6 to recover the Metro map
and type D (for distribution). You'll see 50 counties listed, all coded as M 1, M meaning they are
"metro" counties. Press the down arrow to go to the second screen. You'll see another 48 counties
listed as metro and the last two as non-metro (coded as "0"). (If you continued to down arrow
through the rest of the counties, you would find them all coded as non metro.) Press the home key
to return to the top of the distribution list. You should find all of the counties listed above in the
metro counties.
Another way to look at population is in terms of density, measured as the number of people per
square mile. It stands to reason that the highest population density should be in urban areas.
Compare variable 7 (1990: PERCENT URBAN) to variable 84 (1990: PERCENT PER SQUARE
MILE). Do the maps look the same?
yes
no, not really
Look at the correlation coefficient r in the upper right hand corner of the screen. (To review what
"r" is and how it works refer back to page 16 in Chapter 2.)
The correlation r is
. It is
weak
moderate
strong
(circle one).
Look at the maps closely and describe the areas where population density is the highest.
46
�Migration patterns can tell a lot about an area. People like to move and stay in places where there
are good jobs, schools and services-in sum, where there is a good quality of life. When an economy
weakens and jobs disappear, people move elsewhere in search of a better life. In the 1940s and 50s,
more than three million people left Appalachia for work in northern and Midwestern cities. In the
1980s, some parts of the Appalachian region lost population while other parts gained it. The
MicroCase mapping function shows what areas of the region lost and gained populations. The
scatterplot function provides data on associations between migration and levels of income,
poverty, etc.
Select E. Mapping Variables. At the prompt, type 22 and press <ENTER>. As the map unfolds
you'll notice counties color coded according to percentage of increasing or decreasing population
between 1980 and 1990. The lightest (yellow) clusters indicate areas where population is
decreasing. The darker clusters denote the heaviest population increases. Refer back to the early
map of Appalachia's three regions, either by looking in this manual or using the "Comp" (for
comparison) function and entering 3. The maps indicate that
people are leaving the
(Northern/Central/Southern) Appalachian region.
people are moving into the
(Northern/Central/Southern) region.
Press <ENTER> twice to return to the Percent Net Migration map and Press L (for legend). The
legend shows two extremes in migration. The table should contain the following information
-44.2 <=Val<= -8.9
(79)
-8.8 <=Val<= -5.2
(76)
-5.1
<=Val<= -1.1
(84)
-1.0
<=Val<=
4.2
(78)
4.3 <=Val<= 90.3
(82)
The numbers in parentheses ( ) on the right reflect the numbers of counties in each category. Note
that in the first category there are 79 counties in which somewhere between 44.2 and 8.9 percent
of the people moved out of their county between 1980 and 1990. At the other extreme, there are 82
counties in which somewhere between 4.3 and 90.3 percent of the population moved in. Press
<ENTER> to clear the legend from your screen and press Name and you'll see that Gwinnett
County, GA led the region in migration with a percent of 90.3-it almost doubled its population!
Press the down arrow and you'll see that Dawson, Ga. followed closely with a 85.3% increase.
Press the down arrow again and you'll find another Georgia county—Georgia must be on a lot of
people's minds!
To obtain a full list of migration in and out of all 399 Appalachian counties, press D (for
distribution). You will find Georgia's Gwinnett County followed by:
47
�1)
Gwinnett
6)
2)
7).
3)
8).
4)
9).
5)
10).
Why might people moving to this part of Appalachia. Use the web to find out what makes northern
Gwinnet, Dawson, Cherokee and Pike counties in Georgia so attractive. Summarize your findings
here:
Contrast this list with the counties that lost the most population. Arrow down to the bottom of the
list where Jackson, Alabama sits with a loss of 44.2 percent of its population. Look above Jackson
and note the trend: which three states' Appalachian counties lost the most population?
1)
2)
3)
Why would Georgia gain population while these areas loose it? To answer such questions,
researchers look for relationships (also called associations or correlations) between migration and
other variables. Since common sense suggests that people follow employment opportunity, let's
make a "scatterplot" of the relationship between migration and unemployment. Select F.
Scatterplot from the STATISTICAL ANALYSIS menu. At the prompt for the dependent variable
type 22 and press <ENTER>. At the prompt for the independent variable, type 37 and press
<ENTER> twice. You'll get a graph that should look like this:
48
�Figure 3.1
90.3
Y axis =
%MIGRATION
M
I
G
R
A
T
O
N
-44.2
UNEMP 91
3.4
From 3.4%
X axis = UNEMP 91
25.7
To 25.7%
r = -0.422
This seatterplot should look much better on your computer. To interpret the plot, you need to know
the following terms and concepts:
1) Cases: Each dot represents a case. A case in this data is a county, therefore, there are 399 cases.
Note the dots on the top, left hand side of the graph, indicating high in migration rates and low
unemployment. They must be Gwinnett and Dawson counties in Georgia. Type S (for show case)
and then enter "Gwin." The computer types the rest of the name and flashes its location on the
graph. Press <ENTER> and type Dawson. It should be flashing right up there next to Gwinnett.
What counties have high unemployment? Where would they be located on the graph? They would
be found out along the X axis, on the right hand side of the plot. Notice that the dots on the right
side are also relatively low on the vertical Y axis, suggesting that unemployment is associated with
out migration (the percentage of people leaving a county between 1980 and 1990).
2) Variables: A variable is a characteristic that can change or vary from case to case. The variable
which occurs first, or is suspected to cause change in the second variable is called the independent
variable; the second or impacted variable is the dependent variable. In this example, the
unemployment rate for 1991 (UNEMP 91) would be considered the independent variable and
(%MIGRATION) the dependent variable.
3) The Xaxis. Scatterplots show the relationship of two variables to each other along two axes. The
X axis is the horizontal measure for the first or "independent" variable. In the example in Figure
3.1, each case's score on the variable UNEMP 91 is located along the X axis, from 3.4 percent to
25.7 percent.
49
�4) The Y axis is a vertical measure the second or "dependent" variable. In Figure 3.1, the
dependent variable (%MIGRATION) is charted vertically from -44.2 to +90.3. Each county is
plotted on the graph according to its scores on the independent and dependent variables.
5)Association and correlation: Look for the pattern of dots on the graph. The dots seem to descend
from the left to the right. You can check this by letting the computer draw a straight line closest to
all the dots. Press <ENTER> to leave the "show case" function and return to the scatterplot. Press
L (for line) to draw the "regression line." It slopes down from left to right, indicating a "negative
association between the two variables." When a line slopes down like this from left to right, it
suggests that the higher the score on the independent variable(UNEMP91), the lower the score on
the dependent variable (%MIGRATION). In this case, the dependent variable (%MIGRATION)
ranges from -44.2% to 4-90.3, so the interpretation would be that higher unemployment is associated
with out migration and lower unemployment (or higher employment and job opportunities)
correlates with higher migration into a county. The line or "slope" suggesting a negative association
looks like this:
Figure 3.2
90,3
G
R
A
T
I
O
N
-44.2
3.4
UNEMP 91
25.7
6) Outliers: Note that some dots (cases) stand out away from the cluster of dots in the left of the
scatterplot. Cases that score farthest away from the average are called "outliers." The outliers in
this example will tend to be at the top left and the bottom right of the graph. Press O (for Outlier)
and you will see "Outlier: Dawson, GA" appear under the graph and the dot representing Dawson
flash in red on the screen. As you would expect, it is in the upper left of the graph—higher in
migration is associated with lower unemployment. Now the computer asks if you want to delete this
outlier and tells you that if you do, the correlation "r" will change from -0.422 to -0.440 (meaning
that eliminating this outlier for the data set makes the association between variables even stronger).
Press Y (for Yes!) The computer should tell you that the next outlier is McDowell County, West
Virginia. Notice its location on the lower right hand side of the graph—out migration is associated
with high unemployment. Press Y again and you'll see the county with the highest unemployment,
which is
(county, state). If you eliminate this county, the new r =
. Continue to Press Y and you see a pattern suggesting in migration to Appalachian
Georgia and parts of Pennsylvania and out migration from West Virginia.
50
�7) Slopes: The data suggest that unemployment affects migration. We can hypothesize that
unemployment may also have an association with poverty. Press <ENTER> twice to return to the
SCATTERPLOT menu and type 39 for the dependent variable and 37 for the independent variable.
When the graph appears you'll see that the axis X=UNEMP 91 and Y=%POOR89. Press L (for
Line) and you'll see ^positive slope indicating a positive relation between the two variables: Higher
rates of unemployment are associated with higher rates of poverty; Pearson's r = 0.543. The
scatterplot should look like this:
Figure 3.3
52.13
P
O
O
R
3.98
UNEMP 91
25.7
Figure 3.3 is an example of a positive slope, association and correlation. Figure 3.2 is an example
of a negative slope, association and correlation.
Once more, recreate the original
unemployment/migration graph by typing 22 for the dependent variable and 37 for the independent
variable. When the graph appears, press L (for Line) and again you will see an example of a
negative slope. Incidentally, if the regression line doesn't go up or down, but stays horizontal and
parallel to the x axis, it means there is no association between the two variables. In other words, as
the scores on the independent variable increase, scores on the dependent variable stay the same;
Pearson's r = 0.
Updating migration data: Government agencies such as the Census Bureau and the Appalachian
Regional Commission are continually updating their data files. Shortly before this reprinting, the
ARC published migration data for the 1990-1997 period. Did the trends of the 1980s continue in
the 1990s?
Select E. Mapping Variables. At the prompt, type 94 and press <ENTER>. You will see a map
indicating migration patterns for the 1990-1997 period. Press Comp. The screen will divide in two
and take about 20 seconds to make a smaller map of the Appalachian region. You will be asked for
the "Name or number of variable for comparison." Enter 22. Another map will appear showing the
migration patterns between 1980 and 1990.
Do the maps look similar?
Yes
No
What is the correlation coefficient r =
51
Not Sure
�Press <ENTER> three times to return to the variable prompt and type 94. After the map appears,
press D and write down the ten counties with the highest in-migration in the 1990-1997 period.
Compare your results by looking back a page 24 and seeing if these counties also ranked high on
the 1980-1990 list.
1990- 1997
Rank Position in 1980-90 Migration
1)
2)
3)
4)
5)
6)
7)
8)
9)
10).
Go to the bottom of the migration distribution list and record (from the bottom up) the counties
losing the most population.
399)
398)
397)
396)
395)
394)
393)
392)
391)
390)
52
�What might these counties might have in common the could spur out-migration?
Create a new scatterplot, using the new data for migration and unemployment rate for 1996. Select
F. Scatterplot from the STATISTICAL ANALYSIS menu. At the prompt for the dependent
variable type 94 and press <ENTER>. At the prompt for the independent variable, type 96 and
press <ENTER> twice to bypass the subset request. When the scatterplot appears,press Line and
write in the labels for the X AXIS and the Y AXIS and draw in the regression line:
YAXIS
XAXIS
r=
N:
Prob:
Describe the strength (the correlation coefficient "r") and direction (positive or negative) of the
relationship between unemployment and migration in the 1990-1997 period. Is this correlation
stronger or weaker than the correlation for 1980-1990 (see Figure 3.1 on page 27)?
53
�Review. This chapter introduces a number of new terms. Write out definitions for the following list
in your own words. The glossary at the end of the book can help you, but writing descriptions out
will help you make them part of your own working knowledge.
Case
Unit of analysis
Demographics
Population density
Metropolitan Statistical Area (MSA, CMSA, PMSA)
Migration
Correlation coefficient "r"
Levels (scales) of measurement:
Variable
Independent variable
Dependent variable
Categorical variables:
Continuous variable
Xaxis
Y axis
Strength of association between variables
Positive association between variables
Negative association between variables
54
�You need to review these terms until you can explain them to yourself comfortably—until they
become part of your working vocabulary. They will improve your ability to see relationships
between variables and make choices. For example, if you didn't have strong ties to a particular
place but wanted a good job, where would you go?
Georgia
Eastern Kentucky
West Virginia
(check one)
Use the scatterplot function to discover other variables your findings here, using at least six of the
terms listed above to explain your findings. For example, you could report on the relationship
between migration and the percentage of the labor force involved in manufacturing — or mining.
Now that you see how data can help us understand the relationships between variables such as
unemployment, poverty and migration, you're ready to begin an analysis of Appalachia' s economy.
55
�Northern, Central and Southern Appalachian Regions
56
�V. The Appalachian Economy I: Comparing Subregions
Analysis of Variance (ANOVA) in Group Scores
In the last chapter we explored the association of migration in and out of Appalachian counties with
economic factors such as unemployment. Maps and scatterplots were the statistical tools employed
in the exploration process. We can use another tool called analysis of variance, to compare
migration by the three Appalachian subregions—the North, Central and Southern.
From the STATISTICAL ANALYSIS MENU select C. Analysis of Variance. At the dependent
variable prompt, type 22 (or %migration) and press <ENTER>. At the independent variable
prompt, enter 3 (or "Three Regn," an abbreviation for Three Regions,) andpress <ENTER> twice
(bypassing the request for a subset variable). You will see the following graph:
Figure 4.1
90.3
Y%
M
I
G
R
-44.2
VALUEOF
1
THREE REGN
What you see on the computer monitor should be superior to Figure 4.1. Press X for the x axis
variable label. It tells you that the horizontal axis "Three Regn" (three regions) is broken into three
groups: 1) North, 2) Central and 3) Southern Appalachian subregions. Each dot represents a single
county. The counties are clustered vertically in three columns by subregion and run up or down the
plot according to their in or out migration percentage. The rectangular boxes are indicators of the
spread of variance in individual county scores from their group (region's) average or mean score,
which is marked by the horizontal line in the middle of the box. A taller box means that a group's
scores are more widely spread. The boxes mark the spread of scores one standard deviation above
and below the mean score, which includes approximately the 68 percent of that group's cases
closest to the average score. The dots at the extreme top and bottom of the columns represent
outliers. The line running horizontally across the chart connects the means for each group. Notice
how the line slopes down between Groups 1 and 2 and slopes up between Groups 2 and 3. Since
Group 2 is Central Appalachia, this region must have the lowest mean score. Southern Appalachia
has the highest. Press <ENTER> to continue.
57
�Press M (for means) and you'll find the average migration score for each region in the following
table:
Table 4.1 N: 399 (counties/cases)
Region
Missing (counties/cases):
N (number of counties)
Mean (Average % migration)
North
144
-4.344
Central
84
-7.005
Southern
171
0
4.046
Central Appalachia had the most out migration (-7.005). The Northern region also lost population,
while the Southern gained it. Why would people want to leave Central Appalachia? Might it be a
lack of jobs? Press <ENTER> and, at the dependent variable prompt, type 37)UNEMP 91. Enter
3)Three Regn for the independent variable. Note the line connecting the mean scores for each
group; it seems like an inversion of the previous plot: Central Appalachia has the highest mean
score for unemployment and the Southern region had the lowest. Press M (for Mean) and record
the scores below:
Table 4.2 N: 399 (counties/cases)
Region
Missing (counties/cases):
N (number of counties)
North
84
Southern
Mean (Average % unemployment)
144
Central
171
Describe your results: Which region has the highest unemployment? The lowest?
58
0
�Compare Appalachia's three regions according to the following dependent variables:
1. 32) PERCAP 92 by 3) Three Regions
When the table appears, type Y and "write in the long label for PERCAP 92 here;
Press M (for Mean) and record the scores below:
Region
Mean (PERCAP 92)
North
Central
Southern
Convert the numbers under "Mean (PERCAP92)" into dollars and round off to the nearest dollar
($15,249). Describe your findings, region by region. (This may seem a little tedious, but this is part
of data analysis).
Do one more procedure before leaving this table: press A for ANO VA (analysis of variance). Write
in the numbers for the
p =
22 _
p (probability) =
These figures will explained in more detail at the end of the chapter (page 71). Basically, the F
ratio is a measure of how much of the variation in scores occurs within each region as compared
to between the three regions. The higher the F ratio, the bigger the differences between regions. The
"p value" or probability is an indication of our chances of being wrong in claiming the difference
in scores between the regions is "statistically significant." To be significant, the probability of being
wrong must be less than 1 chance in 20 or .05. The Eta square ("E2") indicates how much of the
variance in the dependent variable (PERCAP 92) is explained by the independent variable "Three
Region."
59
�Go back in time to 1981 and record per capita income in 1981: Enter variable 88) PERCAP 81 by
variable 3) THREE REGN
When the table appears, type Y and write in the long label for PERCAP 81:
Press M (for Mean) and record the scores below:
Region
Mean (PERCAP 81)
North
Central
Southern
Remember to round off and convert to dollars.
Now update your findings to 1995. Try variable 92) PERCAP 95 by 3) Three Regions:
When the table appears, type Y and write in the long label for PERCAP 95:
Press M (for Mean) and record the scores below:
Region
Mean (PERCAP 95)
North
Central
Southern
Describe your results. Can you find any trends from 1981 to 1992 to 1995?
60
�2. 27) H.S.GRAD90 by 3) THREE REGN
When the table appears, type Y and write in the long label here:
Press M (for Mean) and record the scores below:
Region
Mean (H.S. Grad90)
North
Central
Southern
Sometimes decision makers want to know how much more one percentage or rate is over another.
They compute this by subtracting one rate from the other, dividing by whichever is the "base rate"
being used for comparison and then multiplying by 100 to convert the ratio into a percentage. In
this case, if we want to know how much the North's mean education rate is higher than the Central
Region rate, we would compute;
North - Central
Central
41.595-30.039
30.079
11.516
30.079 =0.3828x100 = 38.29%
The North's mean percentage of those 25 and over who have a high school degree is 38.29 percent
higher than of the Central subregion.
This method of portraying differences between rates is commonly used by public health officials.
Practice this technique in each of the exercises in the rest of the chapter.
Write in the numbers for the
F =
E2 =
p (probability) =
61
�3. 41)POORFAM89by3)THREEREGN
When the table appears, type Y and write in the long label here:
Press M (for Mean) and record the scores below:
Region
Mean (POOR FAM89)
North
Central
Southern
Anything interesting? The Central subregion's mean score is
subregion. Describe your results:
percent above that of the North
Write in the numbers for the
F =
E2 =
p (probability) =
62
�4: 53) MANUF.S90 by 3) THREE REGN
When the table appears, type Y and write in the long label;
Press M and record the scores below:
Region
Mean (MANUF.S90)
North
Central
Southern
The North subregion's mean salary in manufacturing is
subregion. Describe your results:
percent above that of the Central
Write in the numbers for the
F =
E2 =
p (probability) =
63
�5: 56) RETAIL $90 by 3) THREE REGN
When the table appears, type Y and "write in the long label:
Enter the mean scores:
Region
Mean (RETAIL $90)
North
Central
Southern
The North subregion's mean salary in retail sales is only
subregion. Describe your results:
. percent above that of the Central
Write in the numbers for the
F =
E2 =
p (probability) =
64
�6: 59) SERVICES$90 by 3) THREE REGN
When the table appears, type Y and "write in the long label:
Enter the mean scores:
Region
Mean (SERVICES$90)
North
Central
Southern
The North subregion's mean salary in services is only
subregion. Describe your results:
percent above that of the Southern
Write in the numbers for the
F =
E2 =
p (probability) =
65
�7: 51) % MINE EM90 by 3) THREE REGN
When the table appears, type Y and "write in the long label:
Enter the mean scores:
Region
Mean (%MINE EM90)
North
Central
Southern
Describe these very interesting results, using percentage differences comparisons between the
Northern and Central regions and the Central and Southern regions.
Write in the numbers for the
F =
E2 =
p (probability) =
66
�8: 54) %MANUF.E90 by 3) THREE REGN
When the table appears, type Y and write in the long label:
Enter the mean scores;
Region
Mean (%MANUF.E90)
North
Central
Southern
The Northern subregion's mean percentage of people employed in manufacturing is
above that of the Central subregion. Describe your results:
Write in the numbers for the
F =
E2 =
p (probability) =
67
percent
�9: 66) FARM VAL87 by 3) THREE REGN
When the table appears, type Y and write in the long label:
Enter the mean scores:
Region
Mean (FARM VAL87)
North
Central
Southern
The Northern subregion's mean farm value is
Describe your results:
percent above that of the Central subregion.
Write in the numbers for the
F =
E2 =
p (probability) =
68
�9: 45) DR.RATE90 by 3) THREE REGN
When the table appears, type Y and write in the long label;
Enter the mean scores:
Region
Mean (DR.RATE90)
North
Central
Southern
The Northern subregion's mean value for active nonfederal physicians per 100,000 is
percent above that of the Central subregion. Describe your results: Are there some outliers in the
Northern region that might exaggerate their rate?
Write in the numbers for the
F =
E2 =
p (probability) =
69
�10: 44) INFANT90-2 by 3) THREE REGN
When the table appears, type Y and write in the long label:
Enter the mean scores:
Region
Mean (INFANT90-2)
North
Central
Southern
How do these results differ from previous findings? What other factors (variables) might explain
this difference?
Write in the numbers for the
F =
E2 =
p (probability) =
70
�In the past ten exercises, we have focused on comparing the mean scores of the three subregions
(categories of the independent variable) for a number of dependent variables (infant mortality rates,
families in poverty in 1989, etc.). Analysis of Variance (ANOVA) provides additional statistical
tools to assess the impact of the categories of the independent variable on the scores of the
dependent variable. If you are using this text in a statistics course, your instructor will explain
ANOVA in detail. Briefly, the analysis of variance procedure indicates how much of the variance
between all the scores (399 scores/counties in our data set) is explained by an independent variable.
For example, how much of the poverty in an Appalachian county can be explained by the fact that
it is located in the Central Appalachian subregion? ANOVA computations lead to an "F ratio," in
which the numerator is a measure of the differences between group (Appalachian subregion)
average (mean) scores and the denominator is a measure of the differences in the case/county scores
within each group. In the form of an equation
differences between group means
_
differences between scores within groups
If the independent variable accounts for or explains much of the variance in scores, the F will be
large: 50/2=25. If the groups means are pretty close to each other and most of the variance is within
the groups, the F will be small: 2/50= 04. Return to variable 41)POOR FAM 89, as an example,
for the dependent variable and 3)Three Region as the independent variable. When you get the bar
and whisker plot, press A (for ANOVA). You will find the F ratio=l 18.152. That's a high score,
which suggests that much of the variance in scores is a result of the location of a county in a
particular subregion. The Eta square ( "E2") indicates that 0,374 (37.4 percent) of the variance is
explained by the independent variable "Three Region."
E2 =
Repeat the procedure for 53)MANUF.$90. F =
What percentage of the variance in manufacturing salaries is explained by a county's location in
a particular region? The answer is E2 multiplied by 100:
Find the F ratio for variable 27)H.S.GRAD90
F=
_
E2 =
How much of the variance in the percentage of people over 25 with a high school diploma
is explained by a county's location in a particular region
Find the F ratio for variable 44)INFANT 90-92 F =
E2 =
How much of the variance in the three year infant mortality rate is explained by a county's
location in a particular region?
In summary, you can see that a county's location in one of Appalachia's subregions makes a
difference in manufacturing salaries, poverty and educational opportunity, but not in infant
mortality rates.
71
�You can continue your comparison of Appalachia's three subregions with tables, which explore
relationships between two categorical variables: Press <ENTER> until you return to the
STATISTICAL ANALYSIS menu and select B. Tabular Statistics. At the prompt for the row
variable type 87)DISTRSDCOLL, which is a measurement of the overall economic condition of
counties as assessed by the Appalachian Regional Commission in 1988. The counties are ranked
from "severely distressed" at the low end to "strong/very strong." For the column variable, enter
3)THREE REGN to break the county economic rankings down by the North, Central and Southern
subregions. Press <ENTER> twice to bypass requests for control variables and subsets. When the
table appears, look carefully at the number of severely distressed counties in the three subregions.
Of the 144 counties in the North region, 13 are severely distressed, as are 26 of 171 Southern
counties. But the Central Appalachia has 50 severely distressed counties out of its total of 84. It's
easier to compare these numbers if you convert them to percentages. Press C (for column
percentages) and enter the results below:
North
9.0%
Central
Southern
Central Appalachia's economic difficulty is obvious. Try one more table. Press <ENTER> to return
to a new request for a row prompt and enter 86)PC%US92col. This is a comparison of each
county's average income with the national average. If a county per capita income scored as 100%
of the U.S. average, it would be the same as the U.S. average. If it scored 125%, it would be 25%
higher than the national average. If it scored 50%, its average (mean) income would be half of the
U.S. average. To find out how Central Appalachia compares to the other two subregions, enter
3)THREE REGN again as the column variable. Press <ENTER> twice and, when the table appears,
type C to obtain the column percentages (which appear in blue). Write in the percentage of
counties in each region with incomes of only 45-60% of the national average.
North
Central
Southern
Describe your results:
72
�In this chapter, you have learned about some differences between the three Appalachian subregions
and about two statistical tools helpful in analyzing those differences: Analysis of Variance
(ANOVA) and tabular statistics. Future chapters will provide more information about these
statistical tools and when to use them. For the moment, concentrate on what you have discovered
about Appalachia by writing a comparative summary of the economic status of its three
subregions. Cite the data you have generated on the previous pages! Compare dependent variables
such as per capita income and mean (average) pay in manufacturing as they differ across the three
subregions. Use the new tools at your disposal: percentages more or less; F ratios, probability
scores for statistical significance and Eta square (E 2 ).
73
�This page intentionally left blank
�VI. The Economy II: Industry and Opportunity
Correlations and Regressions
the
two
in
oo
are
the
are
or
is
22
for
as
2)
and 3)
are
6)CITY,
We
by
are
by
can be
a
and the
has
The
is
or
1)
the
on
3)
are
in the
as
We
and
as in
for
are
that on the
"C" is
for the
the
or
"P1
and
(VII).
and
"I"
are
in
a
"B" is for
and
Comparative data explored in the previous chapter found real differences in income, education and
services between the three Appalachian subregions. Continuous variables such as per capita income
and percentage of people over 25 who have completed high school were compared across three
categories: Northern, Central and Southern Appalachian regions. But there also appeared to be
relationships between the continuous variables: For example, Central Appalachia had the highest
unemployment rates and the lowest high school graduation rates. Might these two continuous
variables be associated with each other?
From the STATISTICAL ANALYSIS menu, select F. Scatterplot Type 37 (UNEMP 91) for the
dependent variable and 27 (H.S.GRAD90) for the independent variable mdpress <ENTER> twice.
After the scatterplot appears,press L to obtain the regression line. Notice that the line is almost flat,
indicating no significant relationship between unemployment and percentage of people completing
high school. Regression lines enable you to quickly assess the strength and direction of an
75
�association between two variables by the steepness of its slope: a horizontal line indicates that as
the independent variable increases nothing happens to the dependent variable—it stays the same.
So the steeper the slope, the stronger the relationship. If the slope goes from the lower left of the
plot to the upper right, it is called a "major diagonal" and the relationship is positive: the higher
the high school graduation rate, the higher the unemployment rate. If the regression line moves
down from the upper left towards the lower right, it is a "minor diagonal" and the association is
negative: the higher the high school graduation rate, the lower the unemployment rate. Continuing
our analysis of the Appalachian economy, we might hypothesize a relationship between mean
county education level and unemployment rates: the higher percentage of people with a high school
diploma, the lower the unemployment rate. Let's see:
Figure 5.1. 1991 County Unemployment Rates by percentage of people over 25 with High School
Degrees.
u
N
E
M
P
9
1
3.4
18.86
r = 0.080
H.S.GRAD90
52.56
Prob. = 0.055
Note the flat regression line-almost horizontal-and that the correlation coefficient V1 is only
0.080. It should be at least 0.10 to suggest even a weak relationship. You may recall from Chapter
II (page 16) statistician William Fox's rough measures for strength of association between
variables.
r > .70
.50 < r < .70
Very strong relationship
Strong relationship
.20 < r < .50 Moderate relationship
. 10 < r < .20 Weak relationship
r <. 10
Negligible relationship
76
�By the
is that
=
5.1? "Prob."
for
the
has
a
a
to
in the
can be
to the
399
set was a
the
in the
we
use the
to
two
in the
is
for us
to
that it
in
is
or
are
to
say that the
in the
is
and will be
in the
in
and
is no
are
out of
or one in 20,
It's
or
and
the
the
two
is no
or "p
the
we say p <
the
the
will say that
that the
for an
(or
to
by
as we
the
to the
=
is less
one in a
all
of a
for a
Create your own scatterplot, this time looking at unemployment in 1980 (38: UNEMP 80) as the
dependent variable and median school years education completed by people in Appalachian
counties over 25 years of age (29: MED.EDUC80) as the independent variable. (The median
education variable was not available for 1990, so I had to move back to the previous census.)
1) Describe the slope and direction of the regression line (positive or negative; steep or
gradual):
2) The Pearson Product-Moment Correlation r =
3) This tells us that the higher the level of education, the higher/lower (circle one) the
unemployment rate.
4) Therefore, the relationship between the two variables is strong/moderate/weak (circle
one), positive/negative (circle one).
5) The probability ("prob.") of error in saying we have a relationship between educational
level and unemployment is
.
You now have the data you need to report your findings. For example, in this case you would report
"there was a moderate, negative correlation between mean county educational level and
unemployment in 1980, r = -.27, p>.001. The higher the educational level, the lower the
unemployment rate." The "p value" or probability indicates that virtually no chance (stated as less
than one chance in a thousand) that our correlation is due to random causes.
77
�Researchers are often asked to suggest relationships between variables before the research is
conducted. A statement about a possible relationship between variables is called a hypothesis
(literally "little thesis"). Thus we can initially hypothesize that "There is a strong, positive
association between per capita income and level of educational attainment in Appalachian
counties." Usually, hypotheses are flipped around to suggest that there is no association between
variables. These are called null hypotheses, positing "null" or no relationship between variables.
In this example we would say, "There is no relationship between per capita income and level of
educational attainment in Appalachian counties." Let's see if we can reject our null hypothesis.
Enter 32)PERCAP 92 as the dependent variable and 29)MED.EDUC80 as the independent variable.
We find that there is/is not (choose one) a
,
relationship between per capita income and median education (r =
p>0.001).
,
We-reject--fail to reject-(choose one) the null hypothesis!
Note that, although the data tells us that p = 0.000, it's better to say there's always a chance of the
correlation coefficient being the result of random factors, even if it's only one in a thousand1/1000-0.001!
In the last chapter, you found that the economic indicator rates for Central Appalachia varied
considerably from the other two subregions. Its economy was marked by more mining and less
manufacturing, lower levels of education and services and higher levels of poverty and
unemployment. With correlation analysis, we can further explore the association between these two
continuous variables. On the SCATTERPLOT menu, type in 37)UNEMP91 for the dependent
variable and 51)%MINE EM90 as the independent variable. I'll duplicate the graph here:
Figure 5.2. 1991 Unemployment Rate and Percentage of Labor Force in Mining.
%MINE EM90
r=0.415
Prob. =0.000
78
N:399
�Notice how most of the cases cluster on the lower left; there is little or no mining in most of
Appalachia's counties. Also note the positive slope and that r = 0.415. We can reject the null
hypothesis of no relationship between mining and unemployment; there is a solid, moderate,
positive association between the two variables.
WAIT JUST A MINUTE! MIGHT ALL THOSE NON-MINING COUNTIES, STACKED UP
LIKE A BAD CASE OF ACNE IN THE LOWER LEFT HAND CORNER OF THE PLOT,
EXAGGERATE THE RELATIONSHIP BETWEEN UNEMPLOYMENT AND THE
PERCENTAGE OF THE WORK FORCE EMPLOYED IN MINING?
Well...hmm. Since we know that most of the mining occurs in Central Appalachia (see page 42),
we can use the subset function to narrow our focus to its 84 counties. Again, on the
SCATTERPLOT menu, type in 37)UNEMP 91 for the dependent variable and 51)%MINE EM90
as the independent variable. When asked for the "name or number of variable 1 for defining subset,"
enter 3. The computer says we have chosen "3) Three Regn," and that its low value is 1 and its high
value is 3. It then asks for the lower limit. Type in 2. The computer says that 2 stands for the Central
region and asks for the upper limit. Type in 2 again and press <ENTER> twice to bypass the second
request for a subset. With the regression line added, your graph should look like this:
Figure 5.3. 1991 Central Appalachian counties' unemployment by percentage of
labor force employed in the mines.
25.7
U
N
E
M
P
9
1
3.4
%MINE EM90
r = 0.401
Prob. = 0.000
79
36.63
n: 84
�What does it show? The Central Appalachian subset of 84 counties has a correlation coefficient
r = 0.401. There still is a moderate, positive relationship saying that higher percentages of workers
engaged in mining is associated with higher levels of unemployment. (And a subset is a group or
groups selected out from a variable containing a larger number of groups.)
Let's continue our analysis by exploring the relationship between mining and poverty. Press
<ENTER> until you return to the STATISTICAL ANALYSIS menu. Select G. Correlation. This
selection gives you the Pearson Product-Moment Correlation r for as many variables as you'd like
to explore at one time. The easiest way to select your variables is to use the F3 key. Press F3 and
use the down arrow to move down the list of variables. Select 51) %MINE EM91 with the left
arrow key, then use the up arrow and left arrow again to select 39)%POOR 89; 41) POOR FAM89
and 43)CHLD POOR89. (The right arrow key provides the long label for each variable.) Press
<ENTER> twice to bypass another the prompt for a fifth variable and the subset request. You will
see a correlation matrix in which each variable is correlated with every other variable.
51)%MINEEM9 39)%POOR89 41)POORFAM89
51)%MINEEM90
1.000
0.478**
0.513**
(399)
(399)
(399)
39)%POOR89
0.478**
1.000
0.986**
(399)
(399)
(399)
41)POORFAM89
0.513**
0.986**
1.000
(399)
(399)
(399)
43)CHLDPOR89
0.473**
0.971**
0.970**
(399)
(399)
(399)
43)CHLDPOR89
0.473**
(399)
0.971**
(399)
0.970**
(399)
1.000
(399)
The first column (columns run up and down; rows left to right) provides the correlations (r) we're
after. Note that r =1.000 in the first cell; that's because %MINE EM90 correlates perfectly with
itself. (A line of 1.000 scores goes diagonally down the matrix, as variables are correlated with
themselves. This is computer junk! Ignore it. Just use the correlations listed below the 1.000 line.)
Note also that the (399) entries are a reflection of the number of cases (399 Appalachian counties).
Now that you see how a matrix works, look at how the percentage of the workforce involved in
mining correlates with the percentage of poor people in a county: r = 0.478. The correlation is 0.513
with poor families and 0.473 with poor children. We can conclude that all of these measures of
poverty have a consistent, strong, positive association with mining in Appalachia.(Remember that
"positive" does not mean "good," but rather tells of the direction of the relationship between the two
variables: as one goes up, so does the other. In a negative relationship, as one variable goes up, the
other goes down. Think of the slope of a regression line in a scatterplot.)
Why should a vigorous, productive activity like mining be associated with unemployment and
poverty? Your instructor can refer you to a number of sources about the history of mining and
corporate activity in the Central Appalachian coal fields. Some authors to explore in an initial
search in your school library or on the Internet are Ronald D Eller, Helen Lewis, John Gaventa and
Steve Fisher. If you live in a coal mining area, you might talk with some miners and coal mine
owners.
80
�Is the relationship of mining to factors such as poverty and unemployment unique? What about
other areas of the economy: manufacturing, retail and services? Construct a correlation matrix for
54)%MANUF.E90,37) UNEM 91,39) POOR 89; 41) POOR FAM and 43) POOR CHILD. List
the correlations below:
54)%MANUF.E90
37) UNEMP 91
r=
39) POOR 89
r=
41)POORFAM89
r=
43) CHLD POOR89
r
Explain these results, remembering that a minus sign (-) indicates a negative (or inverse)
relationship between variables. (An inverse relationship is one where the score on the dependent
variable decreases as the independent variable increases.)
81
�Construct the same matrix again, only this time select Central Appalachia as a subset. At the "enter
a list of variables" prompt, type in 54)%MANUF.E90, 37)UNEM 91, 39)POOR 89, 41)POOR
FAM and 43) POOR CHILD. Press <ENTER> and, at the subset prompt, type 3 and then 2 for both
low and high values. List the correlations below
54)%MANUF.E90
37)UNEMP91
r=
39) POOR 89
r=
41)POORFAM89
r=
43) CHLD POOR89
f* ZSZ
Surprise! I bet you didn't expect this. Compare these results to the previous table:
82
�Construct the same matrix for the North subregion. At the subset prompt, type 3 and then 1 for both
low and high values. List the correlations below:
54)%MANUF.E90
37) UNEMP 91
r=
39) POOR 89
r=
41)POORFAM89
r=
43) CHLD POOR89
r=
Remember the federal government's description of the North subregion as having an old and
outmoded manufacturing base? Manufacturing here is not having the same positive impact it
appears to be having in Central Appalachian counties.
Let's see how poverty associates with different kinds of economies. Variables 50) MINING %90;
52)MANUFCT%90; 55) RETAIL%90 AND 58) SERVICES%90 report the earnings from mining,
manufacturing, retail and services as a percentage of a county's overall income. Some counties will
have much more manufacturing, for example. Does more of one kind of economy associate with
less poverty—do different economic bases carry different economic life chances? At the
CORRELATION menu prompts for variables, type 39, 50, 52, 55, 58. Enter the results in the
following table:
39) %POOR
50) MINING %90
r=
52) MANUFCT%90
r=
55) RETAIL%90
r=
58) SERVICES%90
r=
Mining is positively/negatively (choose one) associated with poverty and the relationship is
weak/moderate/strong (choose one). With r = 0.032, retailing has no association with poverty.
Manufacturing has a strong/moderate/weak (choose one), positive/negative (choose one)
association with poverty, while services have a
(Fill in the blank) relationship. In
summary, it appears that the two economic areas with the strongest correlation with poverty are
and
83
�In this chapter, you have used scatterplots, regression lines and correlations to explore the
relationships between education, types of economies, unemployment and poverty. You have
learned how to use hypotheses and a correlation matrix.
If you were a county commissioner or planner, what kind of economic base would you want for
your constituents? If the software is available, put together a Microsoft PowerPoint or Corel
Presentations slide show, showing the advantages and disadvantages of different economic
development plans. (My students enjoy presenting their plans to each other and other faculty.)
84
�VII: Voting Patterns and Economic Conditions
Tables, x2> and Nonparametric Measures of Association
In previous chapters, we have analyzed relationships between continuous, numerical variables
(frequencies, rates and percentages) and between continuous and categorical variables, such as
3)THREE REGN and 6)CITY. We sought to discover associations among continuous variables with
scatterplots, correlations and regressions. We broke down continuous variables by the values of
a categorical variable, with analysis of variance (ANOVA). Now we will use tabular analysis to
explore relationships between two categorical variables.
Let's begin by constructing some tables showing voting by Appalachian subregion. From the red
STATISTICAL ANALYSIS menu, select B. Tabular Statistics. At the "Enter the name or number
of the row variable" prompt, type 83)PRES.VT92. Press <ENTER>. At the column variable
prompt, enter 3)THREE REGN. Press <ENTER> three times to bypass the request for a control
variable and a subset variable. You will see a table with the following information:
Table 7.1. Counties voting Republican or Democratic in the 1992 presidential
election broken down by the three Appalachian subregions.
Central
North
DEMOCRATIC
66
REPUBLICAN
TOTAL
Southern
53
66
185
76
30
103
209
Missing
2
1
2
5
TOTAL
142
83
169
394
.
The table tells us that of the 142 counties reporting in the North subregion, 66 went Democratic and
76 Republican. There is no data on two counties. For the Central region, 53 went Democratic and
30 Republican and in the Southern region it was 66 and 103. In the TOTAL column on the right,
we see that the Democrats took 185 counties in Appalachia as a whole and the Republicans beat
them with 209. It appears that the Republicans did better in the North and Southern subregions
while the Democrats carried the Central subregion. But this is a table of raw numbers or what
researchers call "frequencies." The problem with frequencies is that it's hard to make comparisons
between the different regions. What does "66" mean when compared to "53?" If frequencies are
converted to percentage ratios, where each column totals 100 (instead of 142-83-169), making
comparisons is easy.
Press Col.% (column percentages). Below the frequencies you'll see the percentages appear, color
coded in blue.
85
�Table 7.2: Frequencies and percentages of counties voting Republican or
Democratic in the 1992 presidential election, broken down by the three
Appalachian subregions.
North
Central
Southern
TOTAL
DEMOCRATIC
66
53
66
185
% Democratic
46.5
63.9
39.1
47.0
REPUBLICAN
76
30
103
209
% Republican
53.5
36.1
60.9
53.0
Missing
2
1
2
5
TOTAL (frequencies)
142
83
169
394
100.0
100.0
100.0
100.0
TOTAL
(Percent)
The percentages immediately enable us to make comparisons. In 1992, the Democrats carried
46.5% of the North counties, 63.9% of the Central and 39.1 % of the Southern. They're stronger
in Central Appalachia, but the Republicans carried 53 percent of the region as a whole. Have these
percentages changed over time? Remember Ronald Reagan? He was elected president in 1980.
How did the Republicans do when he led their party?
Select B. Tabular Statistics. At the "Enter the name or number of the row variable" prompt, type
80)PRES.VT80. Press <ENTER>. At the column variable prompt, type 3)THREE REGN. Press
<ENTER> three times to bypass the request for a control variable and a subset variable. Enter only
the percentages (listed in yellow) below:
Table 7.3: Percentages of counties voting Republican or Democratic in the 1980
presidential election, broken down by the three Appalachian subregions.
Percentage of
counties voting:
North
DEMOCRATIC
Central
Southern
26.1
100.0
100.0
REPUBLICAN
TOTAL
100.0
Comparing the percentages found in Table 7.2 to those in Table 7.3, we find that the subregion that
had the highest percentage change between 1980 and 1992 was the
Northern
Central
— Southern (circle one), with a difference of _
86
_percent.
�Why might voters in this region have changed party allegiance between 1980 and 1992? Did
anything happen to the economy or political life of the nation to prompt this switch? If so, why
might the chosen party policies be preferable, given the change in economic fortunes?
Let's look at the Republicans. Did the voters like Ronald Reagan in 1984, after his first four years
in office? Enter 81 as the row variable and 3 as the column variable and/ill in the cells in Table 7.4.
Table 7.4: Percentages of counties voting Republican or Democratic in the 1984
presidential election, broken down by the three Appalachian subregions.
Percentage of
counties voting:
North
Central
Southern
100
100
100
DEMOCRATIC
REPUBLICAN
TOTAL
Hmm. If I were the president, with these results, I'd be pretty
87
�Are there differences in voting patterns between the three subregions? Remember Central
Appalachia has been dominated by a single industry—the coal industry, with a long history of
conflict between working people and large corporations that control land, resources and often local
government, including law enforcement. It is not surprising that residents would support a political
party with a more activist view of the federal government working to protect the environment and
the rights and well being of blue collar workers. You have also seen in chapters 1 and 4 that Central
Appalachia is, economically, the worst off of the three regions. Do counties in increasingly difficult
economic straits favor the Democrats? Let's hypothesize (make an educated guess of) a positive
correlation between increasing economic distress and support for the Democrats—that higher
percentages of distressed counties will vote Democratic.
The Appalachian Regional Commission's Distressed
Counties Program began in 1983 to provide special funding
for the region's poorest counties. To qualify for Distressed
County status, a county must have an unemployment rate at
least 150% of the U.S. rate of 5.7 (8.6% or higher), 150% of
the U.S. poverty rate of 13.1% (19.7% or higher) and less
than 67% of the U.S. Per Capita Market Income of $19,305
($12,934 or lower) or 200% of poverty and one other
indicator. (Per capita market income is per capita income less
transfer payments.)
At the row variable prompt, type 83)PRES.VT92 and, at the column variable prompt, enter
87)DISTRSDCOL. This last variable ranks counties from severely distressed to economically
strong. Write in the percentage of counties supporting the Democrats in each of the four categories
of the independent variable DISTRSDCOL:
Table 7.5 Counties going Democratic or Republican in the 1992 presidential
election by economic status of county.
Severely
Distressed
Distressed
Middle
Strong
Very Strong
Democratic
Republican
Do you see any trend here? Look across the rows. The more distressed the counties, the more they
tend to favor the Democrats: 65 percent of the severely distressed counties support them, while 70
percent of the counties in strong (good economic) condition voted Republican.
�Notice that the cell differences are mostly greater than 10 percent across the table. Researchers
assume that differences greater than 10 percent are "significant." They also have a more precise
way of testing for significance, called a "chi-square." Press Stat (for statistics) and you'll see that
Chi-square: 15.608 (Prob. = 0.000)
To repeat what has been discussed earlier (see the box on page 29), "Prob." stands for the
probability of error in positing an association between two variables in a population. Probability is
used to determine the chances of being wrong in claiming that a relationship found between
variables in a sample will also be found in the population from with the sample was drawn. If we
were working with a sample taken from all the counties in the United States, our chances of being
wrong in asserting there is a relationship between party choice in the 1992 election and county
economic status throughout the country is less than one in a thousand (p < .001). Those are good
odds at any race track. But here we are not dealing with a sample, but the entire population of
Appalachian counties, and therefore the "p value " indicates the chances that our chi-quare ( x2)
was a product of random factors, rather than being an indication of association between variables.
Still, the odds are a thousand to one supporting our assertion of a relationship between variables.
The probability score must be less than 0.05 (less than one chance in twenty) to indicate
"significant" association between variables. The probability or "p value" is the first thing to look
for to see if a statistical measure (such as x2) is significant. If p <.055 you have significance; if
p > .05 the results are not significant (written as "n.s.").
On the statistics screen below the chi-square and probability measures, you will see V:0.199
C:0.195 and Lambda (DV=83):0.146. These "chi-square based measures of association" are
indicators of the strength of the relationship between the two variables. "C" stands for Pearson's
coefficient of contingency and "V" for Cramer's V. "Lambda" is Guttman's coefficient of
predictability. You can learn more about these measures and how to compute them in a statistics
course, but here you just need to know how to read them. (You don't have to be a mechanic to drive
a car. You can use these indicators just the way you use your car without having to build your own
engine first.) How do you interpret these numbers? Roughly, a score under 0.10 suggests a weak
relationship, scores between 0.10 and 0.30 are moderate and those higher than 0.30 are strong.
All three measures (C, V and Lambda) here suggest a moderate relationship between the counties'
economic condition and party choice.
You have found two independent variables (subregion and county economic status) that affect
voting patterns. Let's look for others. Does location of residence affect voting patterns? Enter 83
as the dependent, row variable and 6)CITY as the independent variable. Press <ENTER> until the
table appears and then press C to obtain column percentages:
What percentage of the nonmetropolitan counties supported the Democrats?
What percentage of the nonmetropolitan counties supported the Republicans?
89
�What percentage of the metropolitan counties supported the Democrats?
What percentage of the metropolitan counties supported the Republicans?
Press R for row percentages.
What percentage of the Democratic supporters were nonmetropolitan?
What percentage of the Democratic supporters were metropolitan?
What percentage of the Republican supporters were nonmetropolitan?
What percentage of the Republican supporters were metropolitan?
Press S for statistics, x2 (chi square) =
C=O
V=O
(prob. = 0.
)
Lambda (DV83)= O
Was there a significant difference between urban and rural voters in the 1992 presidential election?
Yes
No
Check one and use your data to explain your choice below:
What about differences between counties in highland areas (over 1000 feet in altitude) and the rest
of the region? Enter 83 as the row variable and 5 (Highlands) as the column variable. Press
<ENTER> until the table appears and then press "C" to obtain column percentages. After
comparing percentages, press "S" for statistics.
x2=
C=O
.
Prob.=
V=O
Lambda (DV83)=CL
Was there a significant difference between highland and non-highland voters in the 1992
presidential election? Use your data to explain your choice:
90
�Did people in counties with higher mean per capita income vote Republican in 1992? Variable
86)PC%US92COL is the continuous variable 32)PCAP%US broken down into four categories: 1)
45-60% of the average U.S. per capita income in 1992; 2) 61-75%, 3)76-90% and 4)91-125%. Let's
hypothesize that increasing per capita income means increasing percentage of counties favoring
the Republican ticket. Enter 83)PRES. VT92 as the row variable land 86)PC%US92 as the column
variable. When the table appears, press C for column percentages and record the results below:
Table 7.6 Counties going Republican or Democratic in the 1992 presidential
election by percentage of average U.S. per capita income.
45-60%
Democratic
61-75%
76-90%
91-125%
65.5
Republican
Do the percentages suggest a trend here? Describe it:
Even though we're dealing with population data and not a sample of a population, let's see if chisquare and chi-square based measures of association support our hypothesis:
Chi-square (x2) =
C=O
(Prob = 0.006)
V=0.
Lambda (DV83)=(L
Can we accept or reject our hypothesis about a relationship between increasing affluence and voting
Republican?
91
�GOOD! You can now make and read tables and interpret measures of association. Remember
what was said at the start of the chapter about using tables for categorical variables and
scatterplots and correlations for numerical ("interval/ratio") variables. What kind of variable
is 83)PRES.VT92? Press F3, arrow down to 83 and press the right arrow to see its three values:
1) DEMOCRATIC 2) REPUBLICAN 3) PEROT. These values represent three categories of this
categorical variable. Now use the up arrow to 77) DEM.VOTE92, press the right arrow and write
its long label here:
1992:
Notice that, instead of three categories, we have percentages: a range of thousands of possible
scores ranging between 0 and 100 (0.01...26.58...65.21..etc). The data is continuous, not
categorical Each county potentially has a different score from every other county and thus, in a
table, each county would demand a separate row. The table would have too many rows to be of any
use. So, instead of the tables that work so well for categorical data, scatterplots, regressions and
correlations are employed in the analysis of relationships between continuous variables. You have
already seen this in Chapter V. Let's track some voting data using continuous variables. From the
STATISTICAL ANALYSIS menu, select F. Scatterplot
Let's see if the Democrats did better in areas where there was greater poverty: Enter
77)DEM. VOTE92 as the dependent variable and 39)%POOR 89 as the independent variable. Press
<ENTER> twice to bypass the subset request. When the scatterplot appears, press Line, write in
the labels for the X AXIS and the Y AXIS and draw in the regression line:
YAXIS
77)DEM.VOTE92
XAXIS
39)% POOR 89
r=
N:
Prob:
92
�To review methods for interpreting these results, refer back to the opening of Chapter V (pp53-54).
Describe the strength and direction (positive or negative) of the relationship between percentage
of poverty in a county and the percentage of votes for the Democrats:
Did the Democrats do better in counties having higher unemployment. Enter 77)DEM. VOTE92 as
the dependent variable and 37)UNEMP 91 as the independent variable. Press <ENTER> twice to
bypass the subset request. When the scatterplot appears, press Line and write in the labels for the
X AXIS and the Y AXIS and draw in the regression line:
YAXIS
77)DEM.VOTE92
37)UNEMP91
r=
Prob:
XAXIS
N:
Describe the strength and direction (positive or negative) of the relationship between unemployment
and the percentage of votes in a county for the Democrats:
93
�Hmm. I wonder if these results suggest that the higher a county's average per capita income, the
higher the Republican vote. Enter 78)REP. VOTE92 as the dependent variable and 32)PERCAP 92
as the independent variable. Press <ENTER> twice to bypass the subset request. When the
scatterplot appears, press Line, write in the labels and draw in the regression line:
78)REP.VOTE92
X
32)PERCAP 92
r=
N:
Prob:
Describe the strength and direction of the relationship between the counties' average 1992 per
capita incomes and percentages of votes for the Republicans.
94
�It looks like support for the Republicans was fairly steady across income levels. Might that have
changed from the 1980 election? Enter 70)REP.VOTE80 as the dependent variable and
88)PERCAP$81 as the independent variable. Press <ENTER> twice to bypass the subset request.
When the scatterplot appears, press Line and draw in the regression line:
70)REP.VOTE80
X
88)PERCAP$81
r=
N:
Prob:
Describe the strength and direction of the relationship between the counties' average 1981 per
capita incomes and percentages of votes for the Republicans.
Do you see any differences between the 1992 and the 1980 results?
95
�HEY! WHAT ABOUT ROSS PEROT-THE INDEPENDENT PARTY CANDIDATE! WHAT
CORRELATIONS CAN WE FIND WITH HIS CANDIDACY? DID THE POOR CITIZENS,
DEPRIVED OF EDUCATIONAL AND ECONOMIC OPPORTUNITY AND THUS FED UP
WITH "THE SYSTEM" OF THE TRADITIONAL PARTIES, THROW THEIR SUPPORT TO
HIM? WAS IT A MATTER OF "RADICALS FOR ROSS?" Enter 79)PER.VOTE92 as the
dependent variable and 32)PERCAP 92 as the independent variable. Press <ENTER> twice to
bypass the subset request. When the scatterplot appears,press Line and draw in the regression line:
79)PER.VOTE92
X
32)PERCAP 92
r=
N:
Prob:
Describe the strength and direction of the relationship between the counties' average 1992 per
capita incomes and percentages of votes for Ross Perot:
96
�What about support for Perot and levels of educational attainment? Enter 79)PER.VOTE92 as the
dependent variable and 27)H.S.GRAD90 as the independent variable. Press <ENTER> twice to
bypass the subset request. When the scatterplot appears, press Line and draw in the regression line:
Press Xand enter the long label for 27)H.S.GRAD90:
79)PER.VOTE92
X
27) H.S.GRAD90
r=
N:
Prob:
Describe the strength and direction of the relationship between the counties's average 1990 percent
of people 25 and over who have a high school degree and the percentages of their votes cast for
Ross Perot:
Summarize the results of the last two plots. Who voted for Perot, the Independent candidate?
97
�Ross, that's pretty impressive. But with all those better off and better educated people
supporting you, how come you didn't win?
Well, Tom, it doesn't take a rocket scientist to look at the percentages: Return to the DATA AND
FILE MANAGEMENT menu and select E. Codebook. Type 1 and press <ENTER> to send the
data to your monitor screen. You will be asked if you wish to stratify, which means break down one
variable by another as you would in a table. The default setting is N, meaning no. Press <ENTER>
to accept the default setting and again a second time to bypass the request for a subset. Type
79)PER.VOTE92 as the only variable to be listed. You'll see that the mean (average) percentage
of votes for Perot in all the Appalachian counties was only 15 percent. Not a majority-not this time!
Try stratifying this variable by 3)THREE REGN to see if voters in one of the three subregions
might have been more favorable to Perot: Press <ENTER> to begin the codebook process again.
Type 1 and press <ENTER> to send the data to your monitor screen. Type Y at the "Do you wish
to stratify?" prompt. Type 3 as the stratifying variable. Press <ENTER> to bypass the request for
a subset and enter 79 as the only variable to be listed. List the results below:
REGION
N
Mean
North
Central
Southern
Well, how about that for a candidate from Texas! He got about double the percentage of votes in
the
subregion of Appalachia.
In this chapter, you have learned about:
how to use quantitative methods in the analysis of voting behavior
differences in voting behavior in different Appalachian subregions
employing tables in data analysis
the idea of probability
chi-square (x2) as a measure of the ability of a sample's findings to be generalized to a
population
differences between categorical and continuous variables
labels for X and Y axes
regression lines in scatterplots
98
�Review your data and, as you have done previously, write a summary of your findings in this
chapter. If you like, you can gather additional data by looking voting patterns by ethnic group,
location, education and income level. Enjoy!
Another assignment possibility; Write a strategy for a presidential candidate touring the
Appalachian region. What topics should she or he emphasize in different parts of the region?
99
�This page intentionally left blank
�VIII. Race and Region: Minorities in Appalachia
Multiple Regression
In Chapter III you saw that, contrary to the stereotype of an all-Anglo-Saxon Appalachia, the region
is ethnically and culturally diverse, representing many nations and parts of the world. The question
raised in this chapter is whether two of its ethnic groups, Native American and African American,
enjoy the same opportunities and quality of life as other residents.
From the STATISTICAL ANALYSIS menu. Select G. Correlation. Press F3 and use the down
arrow to move down the list of variables and the left arrow to select those listed below. You want
to see how the percentages of Native and African Americans (11:%AMER.IN90) and (18;
%BLACK90) correlate with per capita income (32:PERCAP92), unemployment (37:UNEMP91),
poverty (39:%POOR89); children in poverty (43:CHILD POR89); infant mortality (44:
INFANT90-2); the number of physicians per 100,000 people (45: DR.RATE 90); the percentage
of people over 25 with a high school degree (27; H.S. GRAD90), and the percentage of people with
an Associate Degree or some college (30; COLLEGE90). Write the results below:
Table 6.1 Socio-Economic Indicators by Ethnic Group
11)%AMER.IN90
18)%BLACK90
32) PERCAP 92
37)UNEMP91
39) %POOR 89
43) CHLD POR89
44) INFANT 90-2
45) DR.RATE90
27) H.S. GRAD90
30) COLLEGE 90
List the labels of any correlation r that is greater than plus or minus 0.2:
101
�Compare these results to the guide for explaining strength of association on page 46. Are there
strong associations between ethnic group and opportunity? Describe what you see:
On the whole, discrimination IS/IS NOT (choose one) evident from these data. BUT, WAIT A
MINUTE! Maybe discrimination differs in different parts of Appalachia. We know from chapter
II (page 15) that counties in the Southern subregion tend to have higher percentages of African
Americans than the North or Central subregions. We see just how much more by using the ANOVA
function. From the STATISTICAL ANALYSIS menu select C. Analysis of Variance. At the
dependent variable prompt, type 18j%BLACK90 mdpress <ENTER>. At the independent variable
prompt, type SJThree Regn and press <ENTER> twice to bypass the request for a subset variable.
Notice the steep slope of the mean line on the plot. Type M (for mean) and enter the mean African
American population for each of the three subregions. Repeat the same steps, but this time replace
variable 18)%BLACK with 11)%AMER.IN90.
Table 6.2 Ethnic Group by Subregion
Region
N (number of
counties)
North
84
Southern
Mean
(%AMER.IND)
144
Central
Mean
(% BLACK90)
171
Wow, that's quite a difference! On average, there are more than 61A times as many African
Americans and three times as many Native Americans in the Southern subregion's counties than
there are in the North and Central areas. The Southern region historically has been the one
associated with discrimination against non whites-but it's also more than 30 years after the Equal
Opportunity and Voting Rights Acts and other legislation aimed at promoting equality. Maybe
discrimination and economic inequality have been eliminated. Let's recreate the correlation matrix,
but this time eliminate the 228 northern and central counties which may be concealing some
important relationships and use the Southern subregion's 171 counties as a data subset.
102
�From the STATISTICAL ANALYSIS menu, select G. Correlation. Press F3 and use the left arrow
to select the following variables: 11) %AMER.IN90, 18) %BLACK90); 32) PERCAP92;
37) UNEMP91; 39) %POOR 89, 43) CHILD POR89; 44) INFANT90-2; 45) DR.RATE90;
27) H.S. GRAD90) and 30) COLLEGE90. At the prompt for a subset type 3. Enter 3 for the "lower
limit" and 3 again for the upper limit. Press <ENTER> twice to bypass a second subset request and
wait for the matrix to appear. Write in the results:
Table 6.3 Subset: Southern Appalachian Subregion
11)%AMER.IN90
18)%BLACK90
32) PERCAP 92
37) UNEMP 91
39) %POOR 89
43) CHLD POR89
44) INFANT 90-2
45) DR.RATE90
27) H.S. GRAD90
30) COLLEGE 90
List the labels any correlation r that is greater than plus or minus 0.2 (which therefore has at least
a moderate strength of association) and, in a simple sentence, explain the association between the
two variables:
103
�Since the percentages of Blacks in southern Appalachian counties are so much higher than those
of Native Americans, let's focus our analysis on African Americans. Begin by comparing
correlations found in all of Appalachia with those of the Southern subregion. Fill in the second
column with data from Table 6.1 and the third column with the correlations found in Table 6.3
Table 6.4 Blacks in Appalachia and the Southern Subregion
BLACKS IN ALL
OF APPALACHIA
BLACKS IN
SOUTHERN
SUBREGION
32) PERCAP 92
37)UNEMP91
39) %POOR 89
43) CHLD POR89
44) INFANT 90-2
45) DR.RATE90
27) H.S. GRAD90
30) COLLEGE 90
Describe the results of this table, row by row, taking special note of any large differences. Then
summarize your findings.
104
�Might these results be affected by location? Does living in Southern Appalachian metropolitan
counties including or adjoining cities improve economic opportunities for African Americans?
From the red STATISTICAL ANALYSIS menu. Select G. Correlation. Press F3 and use the left
arrow to select the following variables: 18) %BLACK90; 32) PERCAP 92; 37) UNEMP91; 39)
%POOR89, 43) CHILD POR89; 44) INFANT90-2; 45) DR.RATE90; 27) H.S. GRAD90 and
3Q)COLLEGE90. At the prompt for a subset, type 3. Enter 3 for the "lower limit" and 3 again for
the upper limit. At the prompt for a second subset type 6. Enter 1 for the "lower limit" and 1 again
for the "upper limit." Press <ENTER> twice to bypass a third subset request and wait for the matrix
to appear. Record the results in the second column of Table 6.5. Enter data from the previous
table's (Table 6.4) third column in the third column here:
Table 6.5 Blacks in the Southern Subregion & Southern Metro Areas
BLACKS IN
SOUTHERN
SUBREGION
METRO AREAS
BLACKS IN
SOUTHERN
SUBREGION
(from Table 6.4)
32) PERCAP 92
37) UNEMP 91
39) %POOR 89
43) CHLD POR89
44) INFANT 90-2
45) DR.RATE90
27) H.S. GRAD90
30) COLLEGE 90
Compare the results across the rows (for example the difference in the correlations for per capita
income in 1) metro areas as against 2) the entire Southern subregion.
105
�A statistical method called multiple regression provides another way to explore the impacts of
discrimination. Multiple regression allows the researcher to explore the combined effects of two
or more independent variables on a dependent variable, for example the impacts of being rural and
being Black on being poor. Let's see how several variables related to poverty correlate. From the
STATISTICAL ANALYSIS menu, select G. Correlation and enter the following variables:
8)%RURAL90, 18)%BLACK90, 37)UNEMP91, 39)%POOR 89, and 47)F HEAD/C90 (1990:
percent of households that are female headed with own children with no spouse present). When
asked for a subset enter 3 as the variable and 3 again forboth the lower and upper limits. You should
get a correlation matrix for the 171 counties found in the southern region. Enter the results from
the column headed by 39)% Poor 89:
Table 6.6
39) %POOR 89
8)%RURAL90
18)%BLACK90
37)UNEMP 91
47)F HEAD/C90
There are some very strong correlations here (so strong we run the risk of a statistical problem
called "multicollinarity" where the independent variables correlate too highly). We need to
eliminate the overlap between independent variables (a lot of Blacks live in rural areas and are
unemployed) as they affect the dependent variable 39)%POOR 89 to find out which of the
independent variables have the most impact on poverty.
From the STATISTICAL ANALYSIS menu, select I. Regression and enter 39)%POOR 89 as the
dependent variable. For the independent variables, type 8) %RURAL90518) %BLACK90, and
37)UNEMP 91. Focus on the Southern subregion by entering 3 at the request for a subset and
typing 3 for both the low and high values. You should see:
R-SQ=0.486
Figure 6.1
8)%RURALO)
18)%BIAGK90
37)UNEMP91
BETA -0.244
(r-O.229)
BETA -0.480
(rH5.483)
BETA-0.386
(r =0.539)
106
39)%POORB9
�Notice the R-SQ (R2) in the upper right hand corner. It is a measure of the combined effects of the
three independent variables on the dependent variable %POOR. Put another way, 48.6% of the
variance or difference in poverty rates among the 171 southern counties is explained by these three
variables. The correlation coefficients V, that you wrote into the table on the previous page are
listed below the line linking each independent variable to the dependent variable %POOR. Above
the same line is a new measure called a BETA. This is a standardized beta, indicating how much
the dependent variable would change with a change of one standard deviation in the independent
variable, if the effects of the other independent variables were removed. For example, with the
effects of %RURAL and UNEMPLOYMENT removed, we would predict that an increase of one
standard deviation of %BLACK would result in .460 increase in poverty. Since this is the highest
beta in the graph, it would appear that discrimination against Blacks in these Southern counties is
a primary cause of their impoverishment.
WAIT A MINUTE! MAYBE RACE ISN'T THE ISSUE. WHAT ABOUT THE STATE OF THE
FAMILY AND POVERTY? IN TABLE 6.6, THE CORRELATION BETWEEN % FEMALE
HEADS OF HOUSEHOLDS WITH CHILDREN AND % BLACK WAS VERY STRONG (.844)
AS IT WAS WITH POVERTY (.505). MAYBE THE HIGH BETA SCORE FOR %BLACK IS
SPUMOUS (NOT REAL), MASKING A GENUINE CAUSAL RELATIONSHIP BETWEEN
FAMILY STRUCTURE AND POVERTY.
Put another way, the correlation between %Black and poverty suggests the following causal
relationship:
11)%BLACK90
>39)%POOR89
But maybe the situation is:
11)%BLACK
> 47)F HEAD/C90
>39)%POOR
Is 47)F HE AD/C90 an intervening variable and a genuine cause of poverty? Let's see: Select I.
Regression and enter 39) %POOR as the dependent variable. For the independent variables, type
8) %RURAL 90,18) %BLACK90,37) UNEMP 91 and this time add 41)? HEAD/C90. Focus on
the Southern subregion by entering 3 at the request for a subset and typing 3 for both the low and
high values.
107
�Figure 6.2
R-SO0.544
8)%RURAL90
18)%BUkCK90
BETA =0.345
(r-0.229)
BETA =0.089
(r =0.483)
39)%POORB9
37)UNEMP91
47)F HEAD/C90
BETA -0.331
(r=0.53i)
BETA -0.482
(r =0.505)
The R-SQ has been bumped up from 0.486 to 0.544: our ability to explain changes in the poverty
rate has increased. Furthermore, the relationship between percentage of black people in a county
and poverty now appears to be spurious, the beta having dropped from 0.460 to 0.089. The
percentage of female heads of households with children and no husband present appears to be the
most powerful independent variable affecting poverty. To repeat the language used here: the
correlation between %Black and %Poor appears to be spurious; the correlation between female
heads of households with children and poverty seems genuine.
This is a very controversial finding. How do you explain it? Is single parenting a matter of choice,
culture or a product of historical and social conditions? Most single parents are women. Is this an
indication of discrimination against women in wage and salary structures? Is this an indication of
less buying power for wage earners in general-do more households need the buying power of two
wage earners than was the case ten, twenty or thirty years ago? Maybe the relationship is
something like
11)%BLACK ~>47)F HEAD/C90 —>WOMENS' JOBS/WAGE LEVELS—>39)%POOR
Quantitative analysis often raises more questions than it answers, sending us on a search for more
data. Let's continue our inquiry with the data we have at hand by conducting a multiple regression
for the entire Appalachian region: Select I. Regression and enter 39) %POOR as the dependent
variable. For the independent variables, type 8)%RURAL 90,18) %BLACK90,37)UNEMP91 and
47)F HEAD/C90. DON'T select a subset this time to obtain results for all 399 Appalachian
counties.
How much of the variance in the dependent variable can be explained by the independent variables?
In other words, R2 =
.
108
�Write the BETAs into the following graph
Figure 6.3
R-SQ=
8)%RURAL90
18)%BLACK90
BETA =0.382
BETA"
39)%POOR 89
37)UNEMP 91
47)F HEAD/090
BETA =
BETA-
Describe your results, taking special note of the change in 18)%BLACK90 caused by the addition
of 48)%F HEAD/C90 as an additional independent variable. Compare the results for the entire
Appalachian region to those for the Southern subregion. Is there a difference or are they pretty
much the same?
109
�Let's see if there is a similar pattern for Native Americans living in the Southern subregion: Select
I. Regression and enter 39) %POOR as the dependent variable. For the independent variables, type
8)%RURAL 90, 11)%AMER.IN90, 37)UNEMP91 and this time add 47)F HEAD/C90. Again,
choose the Southern subset (3/3/3). You should get the following results:
Figure 6.4
R-SQ=0.542
8)%RURAL90
11)%AMER.IN90
BETA » 0.352
(r =0.229)
BETA = 0.010
(r =0.184)
39)%POOR 89
37)UNEMP 91
47)F HEAD/CiO
BETA = 0.328
(r =0.539)
BETA = 0.559
(r =0.505)
Which independent variable has the strongest association with poverty?
Which independent variable has the weakest association with poverty?
Reviewing the data, do you think race is the most salient cause, or might it be location (urban or
rural) and/or the kind of household structure?
110
�What about other ethnic groups? Does the nature of the household and location maintain their strong
betas when dealing with "non-minority" Caucasian groups such as those claiming English and
German ancestry? Look at the following graph carefully:
Figure 6.5
R-SCH).540
BETA = -0.002
18)%BLACK90
15)%SCT.IRH90
13)%GERMAN90
12)%ENGLISH90
11)%AMER.IN90
8)%RURAL90
(r =0.483)
BETA = 0.119
(r =-0.270)
BETA = -0.206
(r =-0.492)
BETA = -0.303
(r =-0.554)
39)%POOR89
BETA = 0.072
(r =0.184)
BETA = 0.360
(r =0.229)
BETA = 0.412
47)F HEAD/C90
(r =0.505)
Remember that the independent variables' betas are above the lines. Note that the correlations for
Native and African Americans are reduced to insignificance by the multiple regression procedure.
People claiming Scotch-Irish descent have a weak, positive beta, while those claiming German and
English descent have weak-to-moderate negative betas. Location and household type still have the
highest beta scores.
What about educational opportunity and location? At the dependent variable prompt, enter
30)COLLEGE 90. For independent variables, enter 8)%RURAL and 18)%BLACK. Describe your
results: Is location or race the more serious barrier to attending college?
111
�In this chapter, you have examined some of the circumstances affecting the life chances of Native
Americans and especially African Americans living in Appalachia. You have employed correlation
coefficients and multiple regression techniques in your data analysis. You should know the meaning of
r, beta and R2. Using the F3 function key, you can explore correlations and multiple regressions between
other variables. Relate your findings to Susan Emley Keefe and William Turner's discussions of African
Americans as "a racial minority within a cultural minority," who suffer greater deprivation than Blacks
elsewhere in the United States (Keefe, 1998,150; Turner and Cabbell, 1985, xix).
Review the data and summarize your findings. What have you learned about the impact of
discrimination, location, and wage structures on life chances? What questions have the data not
answered and what new questions have been raised? Conclude your report with suggestions for future
research.
112
�Defining Appalachia
The darker portion of the map marks 179 counties in seven states that both the 1962 Ford study and
the U.S. Congress included in their definitions of the Appalachian region. The lighter portion
indicates the counties added in the legislative process surrounding the passage of the Appalachian
Redevelopment Act of 1965 (which excluded 10 of Ford's counties in the Virginian Blue Ridge).
Ford's definition followed a description of "state economic areas," developed in 1950 by the Bureau
of the Census and the U.S. Department of Agriculture, in which counties having a similar economic
base where grouped together. In the following analysis, the darker area is defined as the "core
counties" of the Appalachian region. The surrounding 220 counties, labeled "added counties,"
combine with the core to make the larger region of 399 counties in 13 states.
�IX. A Region with Many Definitions
Appalachia has been defined in terms of student perceptions, historical and geological
characteristics, unique cultural folkways, economy and problems. One observer recently noted that
the boundaries have been drawn so many times that there can be no "correct" definition of the
region. In 1921, sociologist John C. Campbell described Appalachia as a unique cultural area in
the United States including 256 counties in nine states: Maryland, Virginia, West Virginia,
Kentucky, Tennessee, North and South Carolina, Georgia and Alabama. In 1962, a study of the
region's social-economic needs directed by Thomas R. Ford at the University of Kentucky, used
a 1950s U.S. Department of Commerce designation of 189 counties in seven states (Maryland and
South Carolina were excluded). The Appalachian Redevelopment Act of 1965 created a federallydesignated region based on a definition by state governors and representatives in Congress. This
new region was to include 397 (now 406) counties in 13 states. When the database for this book was
created in 1996, the region was defined as having 399 counties. The four new states added by the
Congress to the Ford and Campbell lists were New York, Pennsylvania, Ohio and Mississippi. (For
an Internet map showing different definitions of the region go to
http://www.unc.edu/-whisnant/appal/maps/Appreg.gif.)
Weak economies were cited as the primary reason for including new counties (nearly doubling the
number of counties) in this new region, which would receive millions of dollars in federal aid. Did
politics explode the size of the region? The more states included, the more votes and chances of
passage of an Appalachian development program in Congress. But, if Congress was concerned
about alleviating poverty and economic distress, should it have targeted its funds on the smaller
area previously cited earlier by regional experts, or did it do the right thing by virtually doubling
the size of Appalachia from 179 counties in nine states to 399 counties in 13 states? With almost
40 years of hindsight, we can reflect upon the wisdom of the 1965 legislative process.
To begin our analysis, let's get a bird's eye view of the core region and the added counties. Select
I. Open, Look, Erase or Copy File from the DATA AND FILE MANAGEMENT menu. Highlight
the APCOUNTY file and press <ENTER>. Switch to the STATISTICAL ANALYSIS menu and
select E. Mapping Variables. At the "variable to be mapped" prompt, type 4. When the map
appears, you'll see that all of New York, Pennsylvania, Ohio and Mississippi are excluded from the
core area as well as most of Alabama and portions of Georgia, South Carolina and North Carolina.
Now look at the overall region in terms of economic strength. The comparative economic positions
of Appalachian counties can be found in two variables: The first is 36) DISTRESSED, where
counties were ranked in 1988 on a five-point scale from "severely distressed" to "very strong"
economies. I collapsed the two highest values "strong" and "very strong" together in variable 87)
DISTRSDCOL, because there were only four "very strong" counties in the region (about 1 %). Make
a map of 87)DISTRSDCOL. Press Legend for clarification of the color coding.
114
�Where do the counties in economic difficulty seem to be clustered?
Where are the relatively few (21) economically strong/very strong counties?
For comparative rankings of the counties, select A. Univariate Statistics. At the prompt for the
variable, enter &7,press <ENTER> twice to bypass the subset and you will see a pie chart showing
the percentages of counties falling into one of the four categories. Press B (for Bar chart). You will
see a large green arrow under the first bar, which is identified in the lower left section of the screen
as "1) Severely Dis(tressed)," with a frequency of 89 counties, which make up 22.3% of the
regions' 399 counties. Use the right arrow key to move to the second category, "Distressed," which
has 50 counties or 12.5 percent. Press Table) and you'll see that the cumulative percentage for
severely distressed and distressed counties is 34.8 percent.
Press <ENTER> to return to the variable prompt and again, type 87. At the subset prompt, enter
4)Appal Core and select 2 (Core County) as both the lower and upper limit. Press <ENTER> to
bypass the second subset request. After the pie chart appears, press B for the bar chart. Notice the
difference? What is the percentage of 1) severely distressed counties now?
What is
the percentage of 2) distressed counties ?
We have eliminated the "added counties," tacked
onto Appalachia by the Congressional legislative process in 1964-65. Press Table and record the
cumulative percentage for severely distressed and distressed counties:
.
Hmm. It seems that about a third of the larger, 399-county Appalachia has economically distressed
counties—but nearly one half of the 179 counties in the smaller, core region are in trouble. Let's
make some bivariate tables to compare these 179 counties to the other 220 counties tacked on in
the 1965 legislative process. From the STATISTICAL ANALYSIS menu, select B. Tabular
Statistics. For the row (or dependent) variable, enter 87)DISTRSDCOL. For the column (or
independent) variable, enter 4)APAL CORE. Press <ENTER> three times to obtain a table of the
counties ranked by economic strength, but broken down into the two "core" and "added" regions.
Press C for column percentages. The results should match table 8.1
115
�Table 8.1. Percentage Appalachian Counties by Economic Status and Subregion
Added Counties
Core Counties
TOTAL
SEVERELY DISTRESSED
15.9%
30.2%
22.3%
DISTRESSED
9.5%
16.2%
12.5%
MIDDLE
67.7%
50.3%
59.9%
STRONG/VERY STRONG
6.8%
3.4%
4.3%
100.0%
100.0%
100.0%
Economic Status
List the percentage of severely distressed counties in the two categories:
Core counties
Added counties
List the cumulative percentage of severely distressed and distressed counties in the two categories:
Core counties
Added counties
Are the two regions in similar economic condition? If not, how are they different?
Incidentally, the core region has
counties, while the added region has
116
counties.
�In 1995, the Appalachian Regional Commission again divided the regions counties into two
categories: distressed and not distressed. (See the box on page 78 for definitions of "distressed").
Return to the STATISTICAL ANALYSIS menu and select B. Tabular Statistics. Enter 90 as the
row variable and 4 as the column variable. Label the table and write in the percentages:
Table 8.2.
Economic
Condition
Added
Counties
Core
Counties
Distressed
Not distressed
Total
The data for 1999 Distressed Counties are found in variable 91. Enter 91 as the row variable and,
again, 4 as the column variable and record the percentages:
Table 8.3.
Economic
Condition
Added
Counties
Core
Counties
Distressed
Not distressed
Total
Review the data from Tables 8.1, 8.2 and 8.3. Did the economic situation in the core counties
improve between 1988 and 1999?
117
�Do people in the two regions make the same amount of money? Table 2 breaks mean (average)
incomes down by core and added counties. Use the codebook function to make some comparisons:
From the DATA FILE AND MANAGEMENT menu select E. Codebook. Select 1 to send your
results to the monitor screen. Enter Y (for yes) at the "Do you wish to stratify?" query. Enter
4)APPAL CORE as the stratifying variable andpress <ENTER> to bypass the subset request. Type
32,33,35 as the "list of variables to be included" andpress <ENTER> to obtain the first of three
tables: (Convert the numbers in the table to rounded-off dollars-for example, $15,109.)
Table 8.4:1992 Counties' mean per capita income by Core and Added Counties
N
MeanS
Added Counties
220
15,109
Core Counties
179
13,899
Subtract the core region's mean income from the "added county" category. The average income
in the core region in 1992 was $
less than the average income of the counties added
to the core by Congress in 1965. Press <ENTER> to obtain the second table of incomes, calculated
as a percent of the U.S. average. Fill in the cells for 1992 below. Press <ENTER> again to obtain
the third table and fill in the cells for 1970.
Table 8.5. Incomes as a percentage of the U.S. national average for 1970 and 1992.
1970 Mean %
1992 Mean %
Added Counties
Core Counties
There seems to be a difference between the two categories of counties. In 1992, people in the
Appalachian "core" counties made
percent of the national average per capita income,
percent less than people in the counties added to the region by the Congress in 1965 (whose
average was
percent). Was there a similar gap 30 years ago when the ARC was founded?
What does the table tell you?
118
�Let's examine the differences between the two (core, added) subregions, using B. Tabular
Statistics in the STATISTICAL ANALYSIS menu. The continuous variable 33)PCAP%US92 is
a comparison of each county's average income with the national average. If a county per capita
income scored 100% of the U.S. average, it would be the same as the U.S. average. If it scored
125%, it would be 25% higher than the national average. If it scored 50%, its average (mean)
income would be half of the U.S. average. Here, 33)PCAP%US92 has been collapsed into a
categorical variable: 86) PC%US92COL. Press F3, press the "End" function key to get to the
bottom of the variable list, arrow up to highlight 86 and press the right arrow key to see the
category labels. The poorest counties, where per capita incomes range from 45 to 60 percent of the
U.S. national average, are coded as 1, Per capita incomes between 61-75 percent of the national
average are coded as 2; 76-90%=3 and 96-125%= 4.
At the prompt for a row variable, enter 86. Type 4 for the column variable and press <ENTER>
twice. Press "C" for column percentages and enter them here;
Table 8.6. Percentage of US mean per capita income of Appalachian counties by
percentage core and added subregions.
Percentage of
US Mean
1
Added Counties
(percent)
Core Counties
(percent)
100
100
45-60%
2 61-75%
3 76-90%
4 91-125%
Total %
It appears that the "core" subregion's percentage of counties in the poorest (45-60%) category is
percent higher than the added counties! As you did in chapter 4 (see page 35),
compute this by using the formula below and multiplying the resulting ratio by 100 to convert it
into a percentage.
Core County - Added County
Added County
119
�Now it's your turn to see if there are other characteristics that distinguish these two Appalachian
subregions from each other: Does the core region have a higher dropout rate than the 220 other
counties? Choose C. Analysis of Variance to compare mean (average) dropout rates. Press F3 to
get a list of the 90 variables in this data file. Press S for search, and type in HS and you'll find that
NOT N HS90 is variable 26. Press the right arrow and you will see the variable described. Write
down that description here:
Press <ENTER> to make the variable label disappear from the screen. You then can mark 26 with
the left arrow and press <ENTER> twice so that "26 " appears as the response to the request for
a dependent variable. Type 4)APPAL CORE as the independent variable,press <ENTER> twice
to bypass the subset request and obtain your graph. Press M to obtain the means and write in a title
for the table together with the results:
Table 8.7
Mean %Not
in High School
Subregion
Added Counties
Core Counties
Using the formula on the previous page, it appears that the "core" subregion's mean percentage of
people not in high school is
percent higher than the added counties.
Is there a difference in the two regions' unemployment rates? Type 31 (the 1991 county
unemployment rates) as the dependent variable and, again, enter 4 as the independent variable.
Press <ENTER> twice to bypass the subset request, review the graph,press M to obtain the means
and write in the title and column labels along with the results:
Table 8.8
Added Counties
Core Counties
It appears that the "core" counties' unemployment rate is
the added counties.
120
% higher than
�Compare the poverty rates in the two regions. Enter 39 as the dependent variable and 4 as the
independent variable. When you see the graph, press Y and write down the "long label" of the
dependent variable here:
Press M to obtain the means and write in the results:
Table 8.9. Mean poverty rates in core and added counties
Subregion
Poverty Rate
Added Counties
Core Counties
It appears that the core counties' poverty rate is
percent higher than the added counties.
What about the percentage of elderly people over 65 living in poverty? Enter 40 as the dependent
variable and 4 as the independent variable. List the "long label" for the dependent variable:
Enter the results below:
Table 8.10. Mean percentage of people over 65 below poverty level
Subregion
Poverty Rate >65
Added Counties
Core Counties
It appears that the core counties' poverty rate for people over 65 is
added counties.
121
percent higher than the
�What about children? Enter 43 as the dependent variable and 4 for the independent variable. Which
region has the greater percentage of kids living below the poverty line?
Enter the results below:
Table 8.11. Mean percentage of children below poverty
Subregion
Children in Poverty
Added Counties
Core Counties
It appears that the core counties' poverty rate for children under 18 is
the added counties.
% higher than
What about medical services? Which region has the most doctors? Enter 45 as the dependent
variable and 4 as the independent variable. Write out the "long label" for 45 (press Y on the graph):
Enter the results below:
Table 8.12. Nonfederal physicians per 100,000
Physician rate
Subregion
Added Counties
Core Counties
It appears that the core counties' number of physicians per 100,000 people is
lower than physician rate found in the added counties.
122
�Which region received more federal assistance in 1990? Enter 68)FED$/CAP90 as the
dependent variable and 4 for the independent variable.
Enter the results below:
Table 8.13. Direct Federal Expenditures and Grants Per Capita
Subregion
Expenditures & Grants
(Rounded off in dollars)
Added Counties
Core Counties
Surprised? The core counties received
more federal assistance on a per capita basis than
did the added counties. Might this be related to the Appalachian Regional Commission's Distressed
Counties Program? Hmm. You're going to have to compare the means of Expenditures and Grants
by Distressed Counties. Enter 68)FED$/CAP90 as the dependent variable and 91 for the
independent variable.
Enter the results below:
Table 8.14. Direct Federal Expenditures and Grants Per Capita by Economic Status
Economic
Status
Expenditures & Grants
(Rounded off in dollars)
Distressed
Not Distressed
On average, the Distressed Counties received $_
more than the other counties.
Stated another way, the Distressed Counties received
than the other counties.
123
percent more in Federal funds
�Students interested in public policy formation may want to research the legislative process involved in
the passage of the Appalachian Redevelopment Act of 1965. Was there a connection between the
number of votes needed in the House and Senate to get the bill through Congress and the mandated
definition of the region? Consider the following data from the Congressional Quarterly summary of the
1965 first session of the 89th Congress (Volume XXI, pages 34-37):
Table 8.15. Numbers of Representatives from Appalachian States - 89th Congress (1965)
Core States
Alabama
Georgia
Kentucky
Maryland
North Carolina
South Carolina
Tennessee
Virginia
West Virginia
Added ARC States
Mississippi
New York
Ohio
Pennsylvania
TOTAL
8
10
7
8
11
6
9
5
41
24
27
97
10
5
TOTAL
74
Total from Appalachian Redevelopment Act States
Total in House of Representatives during 89th Congress
171
435
You can read about the legislative process surrounding the creation of the Appalachian Regional
Commission and its region in the same CQ volume, pages 788-797.
Now consider the question asked at the outset of this chapter: Did the Congress act appropriately in
defining the region as 399 counties in 13 states, or should it have focused on the smaller, "core"
counties that both the United States Congress and the Ford study both considered part of the
Appalachian region? As you have done at the conclusion of previous chapters, use the data you have
generated in this one to develop and explain your argument. Be comparative. Be specific.
124
�X. The Internet: Finding New Data in The Information Age
The data in the preceding chapters is already outdated! Computers have tremendously increased the capacity
for data gathering and analysis, making new and revised information available on a continuous basis. How
can you stay current? One source is the Internet, where you can find data at national, regional, state and
county levels. You can access a website directly, if you know the URL (Universal Resource Locator). For
example, the Appalachian Regional Commission's URL is http://www.arc.gov. You can access sites
indirectly through "search engines" such as Yahoo or Google.
Try the direct approach first: see if you can find the latest Appalachian Regional Commission data on
Distressed Counties. You'll need a browser, such Netscape or Explorer. In the "location" (URL) window
at the top of your browser screen, enter "arc.gov."
Welcome to the Commission's home page. You will see half a dozen icons: About ARC: News, Events and
Publications, etc. Each icon links to other pages. Use your mouse to click on Regional Research and Reports.
Click again on Income Rates in Appalachia, (most recent year recorded). Toward the bottom of the page
you will see a list of states in the Appalachian region. Click on any state for the latest county-by-county data
in that state. This is the URL I used to update the data file for this edition of the book.
What is the latest per capita income for the state of Alabama?
What year is this data for?
What is the latest per capita income for Bibb County, Alabama?
Click on the "back" button on your browser to return to the Regional Research page and select "Population
Change in Appalachia, 1990-1999," (or whatever the most recent year given).
Which state gained the most population between 1990 and 199(9)?
Which state lost population in the same period?
. Hmm. You surprised me!
Write in the data for the following counties:
Population 1999
Gwinnett, Georgia
Broome, New York
Henderson, North Carolina
Harlan, Kentucky
125
Percent gain/loss 1990-1999
�Return to the ARC home page and click on the "Site Index" at the bottom of the page. You can see from the
list of topics that you have plenty of information to write a history of the organization. Of course, this history
would be from the agency's perspective, so you would need to balance your research with other points of
view. You'll need other URLs-those of colleges and universities, think tanks and advocacy groups. This is
where the indirect method of Internet access is helpful. In your browser's "location" (URL) window at the
top of your screen enter "Google.com." Just enter a topic-any topic-and you will be given a number of sites
related to your key word. For your research here, enter "Appalachia" as the subject.
How many URLs did you
find?
.
Find any college or university sites? Yes
No
(circle one)
If yes, what is their connection to Appalachia?
Find any advocacy groups? Yes
No
(circle one)
If yes, what were their concerns about Appalachia? _
Find anything about Appalachian music?
Yes
No
How about Appalachian geology?
Yes
No
Type "Yahoo" in the location window at the top of your screen. When the Yahoo search engine appears on
your screen, enter "Appalachia" in the search window. Some of the sites here will be those found in your
Google search, but you will also find some new ones. How many categories
and how many sites
did Yahoo find for Appalachia?
Type "Alta Vista" in the location window at the top of your screen and you'll be taken to yet another search
engine. Again, enter "Appalachia" in the search window.
How many web pages did Alta Vista find for you?
When you use search engines, you are likely to find new sites for whatever you are researching. As you
found with Alta Vista, the amount of the information sometimes can be overwhelming. You can get lost in
it. The direct access method (knowing the URL you need and accessing it) saves time.
126
�Use the direct method to access another government agency, the Census Bureau. In the location window,
enter *1ittp://www.census.gov."
What is today's estimated population for the United States?
What is today's estimated population for the world?
Under the "PEOPLE" category you will find information on the following subjects
Census 2000 • Estimates •
Projections * Housing • Income *
International • Poverty •
Genealogy
There are also categories for BUSINESS and GEOGRAPHY, where you can access maps. This is great place
to go exploring. Some Census data is not based on the decennial population count, but on samples. Estimates
based on samples require some "wiggle room," called confidence intervals.
on
by
are
is a
is
on
of, not a
the
full
the
for the
or
out
the
in
for
of 41.9
a
the
to
5
17 in
of 72.6
30
a
of the
(29,916)
Try a
in
fall
the
In the
a
as
a
is
1,
as
as
90
for
17 in
5
a
17.0
127
10
�Check out McDowell County, in the coal fields of West Virginia. Click on Select a State on the right-hand
side of your screen. Choose West Virginia from the list. Click on Get State Profile. Select "McDowell" from
the list of counties.
McDowell
W.Va.
What was McDowell's population for the 2000 Census?
What was the median household income (199(7) estimate)?
What was the percentage of people in poverty?
What percentage of children were in poverty?
Let's leave the coalfields and check out Madison County, a rural county outside Asheville in North Carolina.
Madison
N.C.
What was Madison's population for the 2000 Census?
What was the median household income (199(7) estimate)?
What was the percentage of people in poverty?
What percentage of children were in poverty?
Look at an urbanizing Appalachian County outside of Atlanta: Gwinnett, Georgia.
Gwiimeft
Georgia
What was McDowell's population for the 2000 Census?
What was the median household income (199(7) estimate)?
What was the percentage of people in poverty?
What percentage of children were in poverty?
The Census Bureau URL is only the tip of the iceberg for Federal agency data. Try www.fedstats.gov. More
than 70 Federal agencies produce statistics. The Federal Interagency Council on Statistical Policy maintains
this site to provide access to statistics and information produced by these agencies for public use. Click on
"Agencies Listed Alphabetically." Under "E," select the Environmental Protection Agency. Click on air
pollution. Choose the "maps" option and then click again on "Non-Attainment areas" and on "ozone" to see
a U.S. map of areas having unacceptable levels of ozone. List the Appalachian states having unacceptable
levels:
128
�1)
4).
2)
5).
3)
Go back to the Fedstats Statistical Agencies page and select the Mine Safety and Health Administration. Use
the agency's search window (in the upper left hand corner at this writing) and type in "20th Century." Select
the hompage for "A Pictorial Walk Through the 20th Century." It begins with a Glossary of Terms,
describing a
colliery
headframe.
tipple
trip
Good terms to know! Describe the Picture No. 9 on the walk:
Use the button at the bottom of the page to return to the MSHA homepage. Click on the "Fatality
Information" button on the left side of the page. Count the "Fatalgrams" reporting deaths for the current
year. Calculate the percentage of deaths that occurred in Appalachian mines.
Number of fatal accidents
.
Percent Appalachian
.
Percent Appalachian
.
Percent Appalachian
What about 1998?
Number of fatal accidents
What about 1997
Number of fatal accidents
129
�What have you learned about mine safety in American and Appalachia?
I work with a number of health-related agencies and community organizations. They often need county-level
data for planning and grant writing. As is the case with most states, North Carolina has a State Center for
Health Statistics. The center's URL is http://www.schs.state.nc.us/SCHS/. Go there and select Health
Statistics. Select North Carolina Pocket Guide. Select the North Carolina Health Atlas and click on the link
to the County Data Book. Choose "Mortality." You should be provided with a spreadsheet of county-level
data. See if you can find the death rate from heart disease for Buncombe County. Is the county rate higher
or lower than the state rate? By what percentage is it higher or lower? (See Chapter V. as to remember how
to compute the percentage difference.)
Buncombe County
North Carolina rate
The Buncombe County rate is
% above/below the state rate>
You now see that data gathering is also done at the state level. The U.S. Census Bureau provides a link to
state data centers at http://www.census.gov/ftp/pub/sdc/www/
List the data sources in the state that interests you the most, together with their URLs:
As you have seen with your search engines, there are a number of sites dealing with Appalachian culture,
history and resources. Here are some URLs to give you a start in Appalachia:
The Center for Appalachian Studies at Appalachian State University has a good site for links to other
regional resources: http://wwwl.appstate.edu/dept/appstudies/
The Appalachian Studies Association, formed in 1977, is a nonprofit, multidisciplinary organization for
scholars, teachers, regional activists and others whose work centers on the Appalachian region. The
Appalachian Studies Association's mission is to encourage study, advance scholarship, disseminate
information, and enhance communication between Appalachian peoples, their communities, governmental
organizations, and educational institutions. Check it out at http://www.appalachianstudies.org/
The Appalachian College Association (ACA) represents 33 member institutions in the states of Kentucky,
North Carolina, Tennessee, Virginia and West Virginia. Its URL is http://www.acaweb.org/
130
�The Appalachian Center at the University of Kentucky, Lexington is a multi-disciplinary institute created
in 1977 to link University of Kentucky resources with Appalachian communities in programs of Research,
Instruction, and Service. This center's URL is http://www.ukv.edu/RGS/AppalCenter/
Lees McRae College provides resources at http://www.lmc.edu/appstudies/links.htm
Appalshop is a media arts and cultural center located in Whitesburg, Kentucky, in the heart of the Central
Appalachian Coalfields. Appalshop produces and presents work that celebrates the culture and voices the
concerns of people living in the Appalachian Mountains. Appalshop began in 1969 as a War on Poverty
program to train mountain young people in media production skills. Rather than leave the region to find work
in the nation's urban centers, the young people created their own nonprofit media company and began
making films about the culture and social issues of Appalachia. While devoted to a particular place,
Appalshop's work addresses universal concerns. See http://ns.appalshop.org/
The Highlander Center in New Market, TN has been training grassroots leaders for nearly 70 years. Its URL
is http://www.hrec.org. The Highlander library catalog is online. To get to it, go to the Highlander Home
Page, and choose <HIGHLANDER>. This will allow access to the entire library catalog, including books,
audio/visual materials and archives. The Highlander web page has links to other Internet sites of interest to
community activists and researchers. Among them will be links to the catalogs at Wisconsin State Historical
Archives and Tennessee State Library and Archives, both of which house Highlander archives, information
on researching corporations, federal government information, environmental, labor and human rights
activism, and African American issues.
The Mountain Association for Community Economic Development, located in Berea, Kentucky, has
provided technical assistance to grassroots organizations in Kentucky and Central Appalachia since 1976.
It has a business development program for projects with the potential of providing jobs to low income
people, a Democracy Resource Center and an Entrepreneurship Initiative. Since 1994, MACED has
sharpened its focus on sustainable development through the Sustainable Communities Initiative and the
Central Appalachian Sustainable Forestry Program. MACED's URL is http://www.maced.orjg/
Women's Initiative Networking Groups (WINGS) is also located in Berea and is associated with MACED.
It is a nonprofit organization that provides entrepreneurial training, marketing consultation and networking
opportunities to low-and moderate-women interested in starting their own business.
http://www.wingsnet.org/
The Brushy Fork Institute at Berea provides training and support services to mountain communities. The
institute's URL is http://www.berea.edu/BrushvFork/default.HTML
Alan J. Banks, a sociologist at East Kentucky University, has constructed an excellent web page, with links
to a variety of data sources, from local, state and federal agencies to think tanks and advocacy groups. His
URL is http://www.socialscience.eku.edu/Ant/BANKS/classpage.htm
Before he retired from the University of North Carolina at Chapel Hill, David E. Whisnant developed a web
page for his course "Hillbilly Highways: Appalachia and America." The web page includes examples of
student projects (becoming experts on a mountain county through web page construction), historical maps
and links to related topics. The URL is http://www.unc.edu/-whisnant/appal/Svlfal97.htm
131
�Scott M. Pearson at Mars Hill College provides examples of biological research in Appalachian
habitats and the use of GIS mapping in research at http://www.mhc.edu/users/spearson/home.htm. Mars
Hill's Center for Assessment and Research Alliances (CARA) with community groups and agencies can be
found at http://www.mhc.edu/cara/.
Within the mountains are regional web sites, such as the Mountain Area Information Center (MAIN) in
Western North Carolina. Through its URL, http ://www.main.nc .us, MAIN provides information on and links
to resources throughout its service area.
Beyond the region, the "American Studies Web" at Georgetown University provides links on a wide range
of subjects. The URL is http://www.georgetown.edu/crossroads/asw/.
What you have learned about quantitative methods and regional studies in this book should enable you to
make good use of the data and other information available on the Internet. As the URLs above suggest, the
Internet also enables you to network with groups and individuals sharing your interests and concerns. For
those interested in Regional Studies, an Internet search can take you to programs across the United States
and the world.
Good hunting, good luck and, most of all, enjoy your new skills and ability to discover!
132
�Glossary
Appalachia
A mountain region in the Eastern United States, with widely varying definitions.
Appalachian Redevelopment Act of 1965
A law passed by the U. S. Congress, defining the Appalachian region as an area in need
of development and which created the Appalachian Regional Commission (ARC) to
plan and administer federal funds for regional development,
Analysis of Variance (ANOVA).
Bivariate technique for analyzing differences among group means. The independent
variable must be categorical and the dependent variable interval/ratio.
Association
Relationship between two variables.
Bivariate Analysis
Analysis of the relationship between two variables.
Categorical Variable
A variable whose values cluster responses into a group or category, such as gender (male or
female) or shoe size (10, WA, 11, IP/a).
Case
A unit of analysis for a study (in this manual, each county is a case).
Cell Frequency
Number in a table indicating count of scores with a given value or joint values.
Chi Square (x2)
Statistic that compares the actual frequencies in a bivariate table with the frequencies
expected if there is no relationship between the variables. Used for tests of statistical
significance and for some measures of association in tabular analysis.
Codebook
Listing of information about variables in a data set.
Continuous Variable
A variable whose values can fall anywhere within a range, such as a percentage (0-100)
or the length of your right foot
133
�Correlation Coefficient (r)
Measure of association between two interval/ratio variables indicating the strength and
direction of their relationship; summary measure of the extent to which cases are clustered
about the regression line.
Correlation Matrix
An array of correlation coefficients.
Cumulative Percentage
Percentage of scores that have a given value or less.
Data
Records of observations on case variables.
Data File
Data set stored in a form that can be used by a computer.
Demographics
The study of human populations
Density
population per square mile
Dependent Variable
A characteristic of a unit of study that is affected by the categories or characteristics of another
(independent) variable.
Ecological Data or Variables
Aggregate data or variables based on spatial or geographic units such as city districts, states,
or countries.
Genuine Relationship
A causal association that does not appear to be explained by an antecedent variable.
Independent Variable
A variable presumed to have an impact on a dependent variable. Years of education (as an
independent variable) may be associated with income level (the dependent variable).
Level of Significance
Probability that a relationship found in sample data occurs by chance if there is no
relationship in the population.
Mean
Arithmetical average of all scores; the sum of cases divided by the number of cases.
134
�Measure of Association
Statistic summarizing the strength (and sometimes the direction) of a relationship.
Metropolitan Statistical Area
Although the definitions and acronyms have changed from census to census, it is a city of at
least 50,000 people and the surrounding counties with significant economic and commuting
links to it.
Migration
People moving in or out of a area, usually measured in terms often year intervals.
Multivariate Analysis
Simultaneous analysis of data for three or more variables.
Negative Association
Relationship in which higher scores on one variable are associated with lower scores on the
other variable.
Nominal Variable
Variable with values that are unordered categories.
Null Hypothesis
Assumption of no relationship in the population.
Ordinal Variable
Variable with values that can be rank-ordered but that are not measured with a fixed unit of
measurement.
Outlier
Score on interval/ratio variable that is unusually low or high.
Parameter
Characteristic of a population. (See Statistic.)
Percentage
Standardized frequency assuming a total of 100 cases.
Population
Set of cases from which a sample is drawn and to which a researcher wants to generalize.
(See Sample.)
Positive Association
Relationship in which higher scores on one variable are associated with higher scores on
the other variable.
135
�r (Pearson Product-Moment Correlation Coefficient)
Measure of association for the relationship between two interval/ratio variables.
r2 (Coefficient of Determination)
The proportion of variation in a dependent variable explained by an independent variable.
R2 (Coefficient of Multiple Determination)
The proportion of variation in a dependent variable explained by two or more independent
variables.
Regression Line (Least-Squares Line)
Summary line on a scatterplot that minimizes the sum of squares of residuals.
Research Hypothesis
A statement of an expectation about the relationship between variables.
Sample
Set of cases taken from a larger population of cases. (See population.)
Scatterplot
Graphic representation of relationships between interval/ratio variables.
Score
Casefs value on a variable.
Spurious Relationship
A statistical association between variables that is not genuine but instead is due to other
antecedent or intervening variables.
Standard Deviation
Measure of variation in scores; square root of the variance. (See Variance.)
Standardized Variable
Variable whose scores have all been converted to Z-scores.
Statistic
Characteristic of a sample. (See Parameter.)
Statistics
1. Numbers that summarize information. 2. Methods for quantitatively summarizing and
generalizing information. 3. Characteristics of a sample.
136
�Subset
Cases selected for an analysis on the basis of their scores on one or more specified
variables.
Tabular Analysis
Analysis of the associations among variables by comparing percentage distributions.
Type I Error
Rejection of a null hypothesis that is true.
Type II Error
Failure to reject a null hypothesis that is false.
URL
Universal Resource Locator: the "address5 of a web site
Univariate Analysis
Analysis of data concerning only one variable.
V (Cramer's V)
Chi square-based, symmetric measure of association for nominal variables.
Variable
Characteristic or property that differs in value from case to case.
X-Axis
Horizontal axis in a scatterplot, usually used for an independent variable.
Y-Axis
Vertical axis in a scatterplot, usually used for a dependent variable.
137
�This page intentionally left blank
�Codebook for the APCOUNTY File Variables
SHORT LABEL
LONG LABEL
1)NAME
NAME OF COUNTY
2) STATES MAP
MAP CODED TO DIFFERENTIATE STATES
3) THREE REGN
APPALACHIA BY NORTH, CENTRAL AND SOUTHERN
REGIONS
4) APPAL CORE
179 COUNTIES CONSIDERED APPALACHIAN BY FORD
(1962) AND THE FEDERAL GOVERNMENT.
5) HIGHLANDS
FEDERALLY DESIGNATED HIGHLAND COUNTIES
6) CITY
IN MSA, CMSA, PMSA, OR NECMA (1) OTHERWISE (0)
7) %URBAN90
1990: PERCENT URBAN
This variable is used by the following variable: 85) %URb90coll
8) %RURAL90
1990: PERCENT RURAL
9) %URBAN80
PERCENT OF POPULATION WHO ARE URBAN 1980
(CENSUS)
10)%RURAL80
PERCENT OF POPULATION WHO ARE RURAL 1980
(CENSUS)
11)%AMER.IN90
1990: PERCENT AMERICAN INDIAN, ESKIMO, OR ALEUT
12)%ENGLISH90
1990: PERCENT REPORTING ANY ENGLISH ANCESTRY
13)%GERMAN90
1990: PERCENT REPORTING ANY GERMAN ANCESTRY
14) %ITALIAN90
1990: PERCENT REPORTING ANY ITALIAN ANCESTRY
15)%SCT.IRH90
1990: PERCENT REPORTING ANY SCOTCH-IRISH
ANCESTRY
16)%SCOTTSH90
1990: PERCENT REPORTING ANY SCOTTISH ANCESTRY
17)%IRISH90
1990: PERCENT REPORTING ANY IRISH ANCESTRY
18)%BLACK90
1990: PERCENT BLACK
139
�19)%>6490
1990: PERCENT OF POPULATION 65 YEARS OR OLDER
20)%<1890
1990: PERCENT OF POPULATION 17 YEARS OR YOUNGER
21)POPULAT90
1990: RESIDENT POPULATION
22) %MIGR80-89
PERCENT NET MIGRATION (CENSUS BUREAU: INCLUDES
CALCULATIONS FOR UNDERCOUNTS)
23) SEX RAT.90
1990: NUMBER OF MALES PER 100 FEMALES
24) %WIDOWR90
1990: PERCENT OF MALES OVER 15 WHO ARE WIDOWED
25) SME COUN90
1990: PERCENT OF THOSE OVER 5 WHO LIVED IN
DIFFERENT HOUSE IN SAME COUNTY IN 1985
26) NOT N HS90
1990: PERCENT OF THOSE 16 TO 19 WHO ARE NOT IN
HIGH SCHOOL AND HAVE NOT GRADUATED
27) H.S.GRAD90
1990: PERCENT OF THOSE 25 AND OVER WHO HAVE A
HIGH SCHOOL DEGREE (ONLY)
28) H.S.GRAD80
PERCENT OF PERSONS 25 YEARS OLD AND OVER
COMPLETING HIGH SCHOOL ONLY 1980 (CENSUS)
29) MED.EDUC80
YEARS OF SCHOOL COMPLETED: MEDIAN SCHOOL
YEARS COMPLETED BY PERSONS 25 YEARS OLD AND
OVER 1980 (CENSUS)
30) COLLEGE 90
1990: PERCENT OF THOSE 25 OR OVER WHO HAVE
COMPLETED SOME COLLEGE OR ASSOCIATE DEGREE
31)COLLEGE80
PERCENT OF PERSONS 25 YEARS OLD AND OVER
COMPLETING 1-3 YEARS OF COLLEGE 1980 (CENSUS)
32) PERCAP 92
PER CAPITA MONEY INCOME IN 1992
33) PCAP%US92
COUNTRY PER CAPITA IN AS PERCENT OF US PER
CAPITA
This variable is used by the following variable:
86)PC%US92COL
34) PERCAP 70
1970 PERCAPITA INCOME
35) PCAP%US70
1970 PER CAPITA INCOME AS PERCENT OF NATIONAL
AVERAGE
140
�36) DISTRESSED
APPALACHIAN REGIONAL COMMISSION 1988
DESIGNATION OF COUNTY ECONOMY. 1 SEVERELY
DISTRESSED, 2 DISTRESSED, 3 MIDDLE, 4 STRONG, 5
VERY STRONG.
This variable is used by the following variable: 87)
DISTRSDCOL
37) UNEMP 91
1991: CIVILIAN LABOR FORCE UNEMPLOYMENT RATE
38) UNEMP 80
CIVILIAN LABOR FORCE UNEMPLOYMENT RATE (BLS)
1980 (BLS)
39) %POOR 89
1989: PERCENT BELOW POVERTY LEVEL
40) POOR>65 89
1989: PERCENT OF THOSE OVER 65 WHO ARE BELOW
POVERTY LEVEL
41)POORFAM89
1989: PERCENT OF FAMILIES BELOW POVERTY LEVEL
42) POOR FAM80
PERCENT OF FAMILIES IN POVERTY 1980 (CENSUS)
43) CHLD POR89
1989: PERCENT OF CHILDREN UNDER 18 BELOW
POVERTY LEVEL
44) INFANT90-2
INFANT MORTALITY: 3-YEAR RATE -1990-92 (ARC)
45) DR.RATE90
1990: ACTIVE NONFEDERAL PHYSICIANS PER 100,000
POPULATION (AMA *SUBJECT TO COPYRIGHT*)
46) %MALE/CH90
1990: PERCENT OF HOUSEHOLDS THAT ARE MALE
HEADED WITH OWN CHILDREN, NO SPOUSE PRESENT
47) F HEAD/C90
1990: PERCENT OF HOUSEHOLDS THAT ARE FEMALE
HEADED WITH OWN CHILDREN, NO SPOUSE PRESENT
48) AGRICUL%90
1990: PERCENT OF EARNINGS FROM AGRICULTURAL
SERVICES, FORESTRY, FISHERIES AND OTHER (BEA)
49)%AGRI.EM90
1990: PERCENT EMPLOYED IN AGRICULTURE,
FORESTRY, AND FISHERIES
50) MINING %90
1990: PERCENT OF EARNINGS FROM MINING (BEA)
51)%MINEEM90
1990: PERCENT EMPLOYED IN MINING
52) MANUFCT%90
1990: PERCENT OF EARNINGS FROM MANUFACTURING
(BEA)
141
�53)MANUF.$90
1990: AVERAGE PAY IN MANUFACTURING
54) %MANUF.E90
1990: PERCENT EMPLOYED IN MANUFACTURING
55) RETAIL%90
1990: PERCENT OF EARNINGS IN RETAIL TRADE (BEA)
56) RETAIL $90
1990: AVERAGE PAY IN RETAIL TRADE
57) %TRADE E90
1990: PERCENT EMPLOYED IN WHOLESALE AND RETAIL
TRADE
58) SERVCES%90
1990: PERCENT OF EARNINGS FROM SERVICES (BEA)
59) SERVCES$90
1990: AVERAGE PAY IN SERVICES
60) GOVERN.%90
1990: PERCENT OF EARNINGS FROM GOVERNMENT
(BEA)
61)%HLTHEM90
1990: PERCENT EMPLOYED IN PROFESSIONAL AND
RELATED SERVICES IN HEALTH
62) %FARMS 87
1987: FARMLAND AS PERCENT OF TOTAL LAND
63) %FARMS '82
FARMLAND AS A PERCENT OF TOTAL LAND 1982
(CENSUS)
64) FARM SZE87
1987: AVERAGE FARM SIZE
65) FARM SZE82
AVERAGE SIZE OF FARM 1982 (ACRES) (CENSUS)
66) FARM VAL87
1987: AVERAGE VALUE OF FARM LAND AND BUILDINGS
PER FARM
67) FARM VAL82
AVERAGE VALUE OF FARM LAND AND BUILDINGS PER
FARM 1982 (CENSUS)
68) FED$/CAP90
1990: DIRECT FEDERAL EXPENDITURES AND GRANTS
PER CAPITA
69) DEM.VOTE80
1980: PERCENT VOTING DEMOCRATIC (CARTER) FOR
PRESIDENT (ERG 'SUBJECT TO COPYRIGHT*)
70) REP.VOTE80
1980: PERCENT VOTING REPUBLICAN (REAGAN) FOR
PRESIDENT (ERG 'SUBJECT TO COPYRIGHT*)
71)DEM.VOTE84
1984: PERCENT VOTING DEMOCRATIC (MONDALE) FOR
PRESIDENT (ERG *SUBJECTTO COPYRIGHT*)
142
�72) REP.VOTE84
1984: PERCENT VOTING REPUBLICAN (REAGAN) FOR
PRESIDENT (ERG *SUBJECTTO COPYRIGHT*)
73) OTH.VOTE84
1984: PERCENT VOTING FOR A THIRD PARTY CANDIDATE
FOR PRESIDENT (ERC *SUBJECT TO COPYRIGHT*)
74)DEM.VOTE88
1988: PERCENT VOTING DEMOCRATIC (DUKAKIS) FOR
PRESIDENT (ERC *SUBJECT TO COPYRIGHT*)
75) REP.VOTE88
1988: PERCENT VOTING REPUBLICAN (BUSH) FOR
PRESIDENT (ERC *SUBJECT TO COPYRIGHT*)
76) OTH.VOTE88
1988: PERCENT VOTING FOR THIRD PARTY CANDIDATE
FOR PRESIDENT (ERC *SUBJECT TO COPYRIGHT*)
77) DEM.VOTE92
1992: PERCENT VOTING DEMOCRATIC (CLINTON) FOR
PRESIDENT (ERC *SUBJECTTO COPYRIGHT*)
78) REP.VOTE92
1992: PERCENT VOTING REPUBLICAN (BUSH) FOR
PRESIDENT (ERC *SUBJECT TO COPYRIGHT*)
79) PER.VOTE92
1992: PERCENT VOTING FOR PEROT FOR PRESIDENT
(ERC *SUBJECT TO COPYRIGHT*)
80) PRES.VT80
1980: VOTE CAST FOR LEADING PARTY (1=DEMOCRATIC
2=REPUBLICAN) (ERC *SUBJECT TO COPYRIGHT*)
81)PRES.VT84
1984: VOTE CAST FOR LEADING PARTY (1=DEMOCRATIC
2=REPUBLICAN) (ERC *SUBJECT TO COPYRIGHT*)
82) PRES.VT88
1988: VOTE CAST FOR LEADING PARTY (1=DEMOCRATIC
2=REPUBLICAN) (ERC *SUBJECT TO COPYRIGHT*)
83) PRES.VT92
1992: VOTE CAST FOR LEADING PARTY (1=DEMOCRATIC
2=REPUBLICAN 3=PEROT) (ERC *SUBJECT TO
COPYRIGHT*)
84) DENSITY 90
1990: POPULATION PER SQUARE MILE
85) %URB90COLL
PERCENT URBAN COLLAPSED INTO PERCENTAGE
QUINTILES.
86) PC%US92COL
COUNTRY PER CAPITA IN AS PERCENT OF US
PERCAPITA COLLAPSED INTO 4 GROUPS
143
�87) DISTRSDCOL
APPALACHIAN REGIONAL COMMISSION 1988
DESIGNATION OF COUNTY ECONOMY. 1 SEVERELY
DISTRESSED, 2 DISTRESSED, 3 MIDDLE WITH 4 STRONG
AND 5 VERY STRONG COLLAPSED.
88) PER CAP$81
PER CAPITA MONEY INCOME IN 1981 (CENSUS)
89) STATESCODE
REGION'S 13 STATES CODED
90) DISTRESS95
APPALACHIAN REGIONAL COMMISSION-DESIGNATED
ECONOMICALLY DISTRESSED COUNTIES IN 1995.
91)DISTRESS99
ARC-DESIGNATED DISTRESSED COUNTIES 1999
92) PERCAP 95
PER CAPITA INCOME 1995 (CENSUS)
93) PERCAP%US 95
PER CAPITA INCOME AS A PERCENT OF THE U. S. PER
CAPITA INCOME, 1995 (ARC)
94) MIGR90-97
PERCENT NET MIGRATION
95) POP97
POPULATION 1997 (ARC)
96) UNEMP 96
PERCENT UNEMPLOYED, 1996
144
�Readings and Media Resources
A selected bibliography of recently published books on Appalachia:
Couto, Richard A.
1994 An American Challenge: A Report on the Economic Trends and Social Issues in
Appalachia. Dubuque: Kendall/Hunt.
Eller, Ronald D
1982 Miners, Millhands and Mountaineers: the Modernization of the Appalachian
South 1880-1930. Knoxville: University Press of Tennessee.
Ergood, Bruce and Bruce E. Kuhre, editors
1991 Appalachia: Social Context, Past Present and Future. Third Edition. Dubuque:
Kendall/Hunt.
Fisher, Stephen L.
1993 Fighting Back in Appalachia: Traditions of Resistance and Change.
Philadelphia: Temple University Press.
Gaventa, John
1980 Power and Powerlessness: Quiescence and Rebellion in an Appalachian Valley.
Urbana: University of Illinois Press.
Higgs, Robert J. and Ambrose N. Manning
1996 Voices from the Hills: Selected Readings of Southern Appalachia. Dubuque:
Kendall/Hunt.
Higgs, Robert J., Ambrose N. Manning and Jim Wayne Miller
1995 Appalachia Inside Out. Knoxville: University of Tennessee Press.
Hinsdale, Mary Ann, Helen M. Lewis and S. Maxine Waller
1995 It Comes from the People: Community Development and Local Theology.
Philadelphia: Temple University Press.
Philliber, William W. And Clyde B. McCoy, Editors
1981 The Invisible Minority: Urban Appalachians. Lexington: The University Press of
Kentucky.
Crandall A. Shifflett
1991 Coal Towns: Life Work and Culture in Company Towns of Southern
Appalachia, 1880-1960. Knoxville: University Press of Tennessee.
Turner, William H. And Edward J. Cabbell, editors
1985 Blacks in Appalachia. Lexington: University Press of Kentucky.
145
�Suggested Readings and Media Resources by Chapter
Chapter II: Exploring a Region through Quantitative Data
Ergood and Kuhre, Section I. What is Appalachia? Especially 1. the Appalachian
Research Collective's "Why Study Appalachia?," 3. Raitz and Ulack's "Regional
Definitions" and 5. John Campbell's "The Southern Highlands and the Southern
Highlander Defined."
APPALSHOP film: Appalachian Genesis
Chapter III: Cultural Diversity in Appalachia
Ergood and Kuhre: II. Historical Background: 9. John Campbell: "Pioneer Routes of
Travel and Early Settlements."
Turner and Cabbell, chapters 6, 10,11 and 16
Higgs and Manning, pp. 27-76: "The First Inhabitants," and "The Mythical Heritage."
APPALSHOP film: Ourselves and That Promise
Chapter IV: Demographics: Urbanization and Migration
Couto, An American Challenge: Chapters 5 and 7: Population, Work Force and Social
Capital.
Ergood and Kuhre, III. Demographic Characteristics. Note especially Stephen E. White,
"America's Soweto: Population Redistribution in Appalachian Kentucky, 1940-1986."
APPALSHOP film: The Long Journey Home.
Additional film about Appalachian migrants to Cincinnati: The Newcomers
Chapter V: The Appalachian Economy I
Couto, An American Challenge:
Chapter 3: Appalachia in a National and International Economy
Ergood & Kuhre: Section V. The Economy: Especially note 26. Salim Kublawi, "The
Economy of Appalachia in the National Context" and 27: Helen Lewis Fatalism or the
Coal Industry.
Eller, Miners, Millhands and Mountaineers, Chapters 2-4
Gaventa, Chapters 2 and 3
APPALSHOP film strip: Clincho: Story of a Mining Town.
APPALSHOP films: 1) Coalmining Women. 2) Frank Jackson, Coal Miner
Chapter VI: The Appalachian Economy II
Couto, Chapters 4-6: Work and Income
Eller, Miners, Millhands and Mountaineers, Chapters 6,7
Gaventa, chapters 4 and 5
APPALSHOP films: 1) Fast Food Women
2)The Buffalo Creek Flood: An act of Man
3) Mine War on Blackberry Creek
146
�Chapter VII: Voting Patterns and Economic Conditions
Gaventa, chapters 6, 8-10
Ergood & Kuhre, Section VI. Polity, especially Branscome's "What the New Frontier
and Great Society Brought."
APPALSHOP film: The Big Lever: Party Politics in Leslie County, Kentucky.
Chapter VIII: Race and Region: Minorities in Appalachia
Turner and Cabbell provide a rich selection of readings from nearly 20 authors
APPALSHOP film: Mabel Parker Hardison Smith
Chapter IX.: Appalachia: A Land of Many Definiitons
Ergood and Kuhre, I What is Appalachia: 3. Raitz and Ulack, "Regional Definitions."
T.Bruce Ergood, "Toward a definition of Appalachia"
APPALSHOP film: Hard Times in the Country: The Schools
147
�ABOUT
PLAUT is a
faculty of Mars Hill University
for the
Applicant
ACA
Asheville^ North Carolina.
on the
1977 until 2005. He is a
He
a
as a
at Pisgah
Services in
�This page intentionally left blank
�LICENSE AGREEMENT
READ THIS LICENSE AGREEMENT CAREFULLY BEFORE OPENING THE DISKETTE PACKAGE, BY OPENING THIS PACKAGE YOU ACCEPT THE TERMS OF THIS AGREEMENT.
MicroCase® Corporation, hereinafter called the Licensor, grants the purchaser of this software,
hereinafter called the Licensee, the right to use and reproduce the following software: American
Government: An Introduction Using MicroCase in accordance with the following terms and
conditions.
Permitted Uses
+ You may use this software only for educational purposes.
+ You may use the software on any compatible computer, provided the software is used on only
one computer and by one user at a time.
+ You may make a backup copy of the diskette(s).
Prohibited Uses
+ You may not use this software for any purposes other than educational purposes.
+ You may not make copies of the documentation or program disk, except backup copies as
described above.
4 You may not distribute, rent, sub-license or lease the software or documentation.
+ You may not alter, modify, or adapt the software or documentation, including, but not limited to,
translating, decompiling, disassembling, or creating derivative works.
+ You may not use the software on a network, file server, or virtual disk.
THIS AGREEMENT IS EFFECTIVE UNTIL TERMINATED. IT WILL TERMINATE IF LICENSEE
FAILS TO COMPLY WITH ANY TERM OR CONDITION OF THIS AGREEMENT. LICENSEE MAY
TERMINATE IT AT ANY OTHER TIME BY DESTROYING THE SOFTWARE TOGETHER WITH
ALL COPIES. IF THIS AGREEMENT IS TERMINATED BY LICENSOR, LICENSEE AGREES
EITHER TO DESTROY OR RETURN THE ORIGINAL AND ALL EXISTING COPIES OF THE
SOFTWARE TO THE LICENSOR WITHIN FIVE (5) DAYS AFTER RECEIVING NOTICE OF TERMINATION FROM THE LICENSOR.
MicroCase Corporation retains all rights not expressly granted in this License Agreement. Nothing
in the License Agreement constitutes a waiver of MicroCase Corporation's rights under the U.S.
copyright laws or any other Federal or State Law.
Should you have any questions concerning this Agreement, you may contact MicroCase
Corporation by writing to: MicroCase Corporation, 1301 120th Avenue N.E., Bellevue, WA 98005,
ATTN: College Publishing Division.
�
Dublin Core
The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/.
Title
A name given to the resource
Appalachian Consortium Press Publications
Description
An account of the resource
This collection contains digitized monographs and collections from the Appalachian Consortium Press.
Creator
An entity primarily responsible for making the resource
Appalachian Consortium Press
Publisher
An entity responsible for making the resource available
Appalachian Consortium Press
Date Issued
Date of formal issuance (e.g., publication) of the resource.
June 1, 2017
Contributor
An entity responsible for making contributions to the resource
<a title="Digital Scholarship and Initiatives" href="http://library.appstate.edu/services/digital-scholarship-and-initiatives" target="_blank">Digital Scholarship and Initiatives</a>
Publication
Digital Publisher
Digital Republication
Appalachian State University
Dublin Core
The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/.
Title
A name given to the resource
People, Politics and Economic Life: Exploring Appalachia with Quantitative Methods
Description
An account of the resource
This is the second edition, published in 1996, of a workbook designed as a supplementary text for courses on regional studies, geography, history, and social sciences. The text teaches students how to explore issues people face in Appalachia. Data from the U.S. Bureau of the Census, Bureau of Economic Analysis, Appalachian Regional Commission, and other agencies that were originally collected for the first edition, were updated. A section was added on how to use the internet to research and obtain information from the national, regional, and county levels. Dr. Susan E. Keefe contributed an overview of the Appalachian region as the first chapter, titled: "Appalachia and Its People".<br /><br /><a href="https://drive.google.com/open?id=13iMlkkg6kKDrUo9o0LcdxioDAtIQheSz" target="_blank" rel="noopener">Download EPub<br /><br /></a><a title="UNC Press Link" href="https://www.uncpress.org/book/9781469641348/people-politics-and-economic-life" target="_blank" rel="noopener">UNC Press Print on Demand</a>
Subject
The topic of the resource
Appalachian Region--Economic conditions
Appalachian Region--Social conditions
Appalachian Region--Population
Creator
An entity primarily responsible for making the resource
Plaut, Thomas
Date
A point or period of time associated with an event in the lifecycle of the resource
1996
Publisher
An entity responsible for making the resource available
Kendall/Hunt Publishing Company
Appalachian Consortium Press
Alternative Title
An alternative name for the resource. The distinction between titles and alternative titles is application-specific.
People, Politics and Economic Life: Exploring Appalachia with Qualitative Methods
Language
A language of the resource
English
Format
The file format, physical medium, or dimensions of the resource
PDF
Textbooks
E-books
Type
The nature or genre of the resource
Text
Rights
Information about rights held in and over the resource
<a href="https://creativecommons.org/licenses/by/4.0/deed" target="_blank" rel="noopener">https://creativecommons.org/licenses/by/4.0/deed</a>
Spatial Coverage
Spatial characteristics of the resource.
https://www.geonames.org/12212302/appalachia.html
Source
A related resource from which the described resource is derived
<a title="UA 76 Appalachian Consortium records" href="https://appstate-speccoll.lyrasistechnology.org/repositories/2/resources/9" target="_blank" rel="noreferrer noopener"> UA 76 Appalachian Consortium records </a>
Is Part Of
A related resource in which the described resource is physically or logically included.
<a title="Appalachian Consortium Press Publications" href="https://omeka.library.appstate.edu/collections/show/82" target="_blank"> Appalachian Consortium Press Publications</a>
Appalachia
demographic
socio-economic
state data