. . .9...
:a‘umfrf

...

,6...
tr

1 alﬁ
hm.”-

. r
.. u.
;

i.

. ......hh».
I .

? Janna!

l.

. v, ...:

c2. .

4.6%"? r

3 .J . as:

us. , k :33
3‘

II

I mug;
m“. .

WWW 3%....

w

35“

rd.

.eﬁﬁ.

 

 

, 3
1“:‘\v
If. ...: .

$52.... f

h
Pb... ..

in.
v p

5‘

 

0’

1. ~
...}... ....a;
7 . .0 1!.‘. 1.5% I
Lip... ”N3: .
ma Faummsulqa .
_M 5.3.9531: 3;. .
. .. aﬁijhanﬂnuuuﬂvlﬂ.
; u....?.f.$
mmuﬂﬁau
I «HAWK
a «....9.
.... .J van
, .u . u... A.—
.. I; l...m.§....l? . ...
, , . nu .. .
3:711 .91., :7 . hwy Jun»- 3
._ . = . {1 ﬂ,
5.04133 we," L E sq . . ‘

o. :1 , .
. ‘ a.
#24.“ 00.“

I.

A!

a)
. :
‘W‘N—x
...». .ﬁww .

1.
«whore?
., ..4.

.s:
O .tcl In.
a

Wﬁm

#130“,

l

 

 

“Ems LIBRARY
3006 Michigan State
University

 

 

 

This is to certify that the
dissertation entitled

MEASURING MSW STUDENT ATTITUDES TOWARD
POVERTY AND THE POOR: A TOOL FOR EVALUATING
SOCIAL WORK EDUCATION OUTCOMES

presented by

MARGARET HELEN WHALEN

has been accepted towards fulﬁllment
of the requirements for the

PhD degree in Social science and social work

 

 

A. a. AA

Major Professor’s Signature

5/3/05

 

Date

MSU is an Afﬁrmative Action/Equal Opportunity Institution

_ .-.—..--.—.n....—-.-0—-_o-o--n—u—o—-----—o-o-..

- -t-~-_.—.—-—-o—.-- — ..

-.-._.—.-—-.—n-» -

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

m

MEASURING MSW STUDENT ATTITUDES TOWARD POVERTY AND THE
POOR: A TOOL FOR EVALUATING SOCIAL WORK EDUCATION OUTCOMES

By

Margaret Helen Whalen

A DISSERTATION
Submitted to
Michigan State University
In partial fulﬁllment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
School of Social Work

2005

ABSTRACT

MEASURING MSW STUDENT ATTITUDES TOWARD POVERTY AND THE
POOR: A TOOL FOR EVALUATING SOCIAL WORK EDUCATION OUTCOMES

By
Margaret Helen Whalen

This study replicates, reﬁnes, and offers an enhanced tool for assessing social
work students’ attitudes toward poverty and the poor. The research concerned ways
social work educators can assess the effectiveness of social work education with respect
to students’ values, prior to or after, professional education.

Seventy, ﬁrst year MSW students, enrolled in their ﬁrst semester course on
organizational and community theories, agreed to participate during the twelfth week of a
ﬁfteen week semester. Students voluntarily, and anonymously, completed a
questionnaire containing the original thirty-seven item Atherton, et a1. (1993) Attitudes
Toward Poverty and the Poor scale (ATP). The questionnaire included seventeen
additional attitude items and twenty-one questions regarding respondents’ demographic
characteristics, personal backgrounds and professional interests.

Factor analysis of the ﬁfty-four item version of the ATP showed no clear factor
structure. Four, internally reliable conceptual subscales were crafted and analyzed with
respect to respondents’ demographics, backgrounds and interests. No associations were
found between total scale score and subscales when correlated with demographic
characteristics or personal backgrounds. Relatively strong correlations were found
among attitude subscales and self-reported inﬂuence of liberal/conservative worldviews

and inﬂuence of religion on attitudes toward the poor and decisions to enter social work.

Respondents with more conservative worldview ratings reported stronger inﬂuence
ratings for extent to which religion inﬂuenced their attitudes toward poverty and the poor.
Respondents with more liberal worldview ratings reported stronger inﬂuence ratings for
extent to which worldview inﬂuenced their decisions to become social workers.

The study demonstrated the viability of the modiﬁed version of the ATP scale,
identiﬁed sound conceptual subscales, and suggested an evaluation logic model for using
the measure for evaluation of professional education with respect to social work values

and students’ attitudes.

Copyright by
MARGARET HELEN WHALEN
2005

DEDICATION

To Bev,
for “only two more years”

that lasted for seven.

ACKNOWLEDGEMENTS

You will hear people talk about education as a journey, but it’s not a journey that
just one person undertakes; the journey often starts with people who are not there at the
end of the trip. My journey, started several generations ago, with an immigrant and an
orphan, my grandmothers. A proud, 17-year-old, Irish Catholic girl, who left Northern
Ireland to work in a Philadelphia garment factory. And a 10-year-old Italian girl, whose
father and family couldn’t care for her and her siblings. Over Tetley tea and orange
marmalade on mufﬁns my Irish grandmother taught me to hate oppression, to ﬁght hard
for justice, and to ﬁnd wealth in small places. My Italian grandmother, and her siblings
were raised in a Catholic orphanage in Philadelphia. From them I learned about making a
family with those you meet along the way; that family is whomever you consider family;
that simple generosity, kindness and responsibility is something you live everyday; and
that you don’t need to have a blood tie to someone to care about them or care for them.

My grandmothers shaped my values and gave me my parents who taught me
about having roots and wings. From my father I learned my history, the importance of
unions, that intelligence and education are two different things and come in many forms.
My father taught me reverence for children, animals, and people who help the small and
weak. My mother taught me the power of having a vision and using my voice. I grew up
not only knowing “You CAN ﬁght city Hall”, but believing I had an obligation to do so.
When I left to get educated I took with me these lessons and the values they had instilled.

I learned a great deal at the colleges and universities I attended. But I never
forget where I really started my Social Work education, long ago, as a child. There are
many people no longer with me at the end of this journey. My grandmothers and great

aunts are gone, my mother, my brother Thomas, but I’m ﬁnishing the journey they

vi

started. Now, at the end of the educational journey, I have many people to thank, my
father Marty and siblings Michael, Patrice, Donna, Maria, Patty, and Marty. To my ﬁve-
year-old brother, Michael, now nearly thirty, I can ﬁnally say, “Yes Mike, I’m a doctor
now!” The faculty I have worked with most closely, Rena Harold, Gary Anderson, Joyce
Ladenson, and Diane Levande, started me along the ﬁnal leg of the journey. Cris
Sullivan, Jan Bokemeier, and John Seita each played a special role in helping me reach
my doctoral destination. Likewise, my friends gave me the encouragement to ﬁnish this
study, most especially Judy, Noriko, DeBrenna, Linda, Patricia, and Rose.

I offer special thanks to the school of social work staff, Mary Lou Jenks, Nancy
Fair, Nancy Gray, Debbie Vasicek, Jacqueline Humphreys, and Carole Abel. They
probably don’t know that I consider them my “everyday family” at the School. For
nearly ten years, they were a soothing reminder to me of where I came from. Each
provided gifts of support, knowledge, insight, forgiveness, praise, or encouragement. My
education would be seriously incomplete without the lessons they shared. I also want to
acknowledge a unique group of supporters, the weekday morning crew at the Flapjack
Shack. Watching me work every weekday morning from April 2004 to today, they
helped me feel grounded, proud of my task, and humbled about my education. Their
early morning gentleness, curiosity, and encouragement sustained me over the months.

Dissertations are made of persistence and encouragement, without both projects
do not end. My partner Bev has been a source of each. Her playful humor, unending
love, and unshakeable belief in me nurtured my mind and sustained my spirit. I have

more love and gratitude than words can say.

vii

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................ x
LIST OF FIGURES ....................................................................................................... xi
CHAPTER 1
INTRODUCTION ...................................................................................................... 1
Professional Social Work Education ................................................................... 4
Social Work History, Poverty, and the Poor ........................................................ 6
Evaluating Social Work Education Outcomes ................................................... 10
Evidence-Based Practice ........................................................................ 11
Effective Social Work Education ........................................................... 11
The Nature of Evidence of Educational Effectiveness ...................................... 12
Target Population ............................................................................................... 14
Evidence-Based Education .................................................................... 14
A Logic Model for Evaluating Social Work Education ......................... 15
Inputs......... ..................................................................................................... 18
Outputs.......... ................................................................................................. 19
Outcomes .................................................................................................. 19
Overview of Logic Model for Evidence-Based Education .................... 19
Student Values as Evidence of Effective Educational Practice ............. 20
CHAPTER 2
REVIEW OF THE LITERATURE .............................................................................. 22
Knowledge, Values, Attitudes and Behavior ..................................................... 22
Research on Social Work Student Attitudes Toward Poverty and the Poor ...... 24
Overview and Critique of the Research on Students’ Poverty Attitudes ........... 33
Socio-Demographic Correlates ............................................................. 35
Purpose of Study ................................................................................................ 36
Research Questions ................................................................................ 37
Summary .................................................................................................. 39
CHAPTER 3
METHODOLOGY ....................................................................................................... 40
Sample Selection and Characteristics ................................................................ 40
Procedures .................................................................................................... 40
Measures ................................................................................................... 42
Attitudes Toward Poverty and the Poor ................................................. 42
Personal Characteristics and Background .............................................. 45
Data Analysis .................................................................................................... 47

viii

CHAPTER 4

RESULTS ..................................................................................................................... 49
Attitudes Toward Poverty and the Poor Survey ................................................ 49
Individual ATP Scale Item Analysis ...................................................... 49
Scale score creation and descriptive statistics for the original
ATP scale ......................................................................................... 53
ATP scale reliability and validity .......................................................... 54
Factor analysis ....................................................................................... 55
Content analysis of 54 items for full scale for subscale creation ........... 56
Poverty beliefs ....................................................................................... 57
Veriﬁable opinions ................................................................................. 58
Policy attitudes ...................................................................................... 60
Personal taxation preferences ............................................................... 6O
Inter-subscale associations ..................................................................... 61
Analysis of socio-demographics, background and individual characteristics 64
Dimensions of social inequality ............................................................. 66
Personal history of receiving government assistance ............................ 72
Interests within professional social work ............................................... 78
Political orientation or world view ........................................................ 79
Religious/spiritual afﬁliation ................................................................. 82
Geographic effects ................................................................................. 84
Findings overview ................................................................................. 84
CHAPTER 5
SUMMARY, CONCLUSIONS, AND IMPLICATIONS ............................................ 86
Summary and discussion ................................................................................... 86
Replication ............................................................................................ 87
Socio-demographics, personal characteristics and
background variables ...................................................................... 9O
Implications for social work ............................................................................. 95
A logic model for evaluating social work education for values
outcomes .......................................................................................... 95
Study limitations ............................................................................................... 96
Directions for future research ........................................................................... 99
APPENDIX ................................................................................................................. 102
Appendix A — Informed consent and study disclosure form
and survey instrument .......................................................................... 103
REFERENCES ........................................................................................................... 111

ix

LIST OF TABLES

Table 1 — Attitudes toward poverty and the poor, original and new scale items. ........... 43
Table 2 - Descriptive statistics for individual ATP items. ............................................. 50
Table 3 — Poverty beliefs subscale items and item-total correlations. ............................ 58 A
Table 4 -— Veriﬁable opinions subscale items and item-total correlations. ...................... 59
Table 5 — Policy attitudes subscale items and item-total correlations. ........................... 60
Table 6 — Personal taxation preferences subscale items and item-total correlations. ..... 61

Table 7 — Pearson correlation coefﬁcients & signiﬁcance levels for ATP subscales. 62

Table 8 — Descriptive statistics for the original ATP scale and four new subscales. ..... 63
Table 9 - Research questions related to each of six major conceptual variables. .......... 65
Table 10 - Respondent history of participation in various social welfare programs. ...... 73

Table 11 — Program recipient differences for scale items related to speciﬁc programs. . 77

Table 12 — Type and extent of student interest in working with or for the poor. ............. 78
Table 13 - Findings for non-directional research questions. ........................................... 90
Table 14 — Findings for directional research questions. .................................................. 91
Table 15 — Findings for differences between recipients and non-recipients. .................. 92

LIST OF FIGURES

Figure 1 — Social work education logic model ................................................................ 17

xi

CHAPTER 1
INTRODUCTION

Growing public demand for evidence;based practice has reached the academic
world of social work educators. Evidence-based pedagogy in professional education
requires evaluation of educational outcomes. Evaluating outcomes of professional social
work education is a task best met by using valid and reliable indicators of knowledge,
skills, and values. Examining changes in social work students’ values is one important
aspect of professional education. Values are often reﬂected in our attitudes. Assessing
students’ attitudes and attitude changes is important for evaluating social work education
outcomes.

Over the last three decades, social service agencies have been challenged to
demonstrate the effectiveness of programs they offer. Government ofﬁcials,
communities, taxpayers, philanthropists and consumers increasingly require
organizational recipients of funding to evaluate the effects of their services they offer
(Barlow, Hayes, & Nelson, 1984; Peebles, 2000). In response there has been a steady
expansion of systematic accountability required of human service delivery systems,
agencies, and programs (Thyer, Isaac, & Larkin, 1997). Rossi, Freeman and Lipsey
(1999) explain,

Evaluations are undertaken for a variety of reasons: to judge the worth of

ongoing programs and to estimate the usefulness of attempts to improve

them; to assess the utility of innovative programs and initiatives; to

increase the effectiveness of program management and administration; and

to satisfy the accountability requirements of program sponsors (p. 13).

Social service providers, evaluation researchers, public and private agencies for

health-oriented, educational, and public welfare programs have reacted to accountability

requirements by offering routine evaluations of both the process and outcome of

delivered services. The social work profession has responded to the growing mandatory
accountability within social service systems by conducting program evaluations (Proctor,
1990; Rubin & Babbie, 2001; Zerhouni, 2003) and teaching social work students to
identify, demonstrate, and provide evidence-based practices (Bloom & Fischer, 1982;
Bloom, Fischer, & Orme, 2003; Garnbrill, 1999; Garnbrill, 2000; Glicken, 2005; Neuman
& Krueger 2003).

In addition to these efforts to evaluate social work practice, social work education
has an ongoing need for systematically evaluating itself with respect to the profession’s
mission and the goals of professional social work education. Self-awareness and practice
accountability are skills regularly taught to social work students. Reﬂection, or self-
study, is an initial step toward professional self-development. Self-study also can be a
preliminary step for professional schools toward cultivating a collective professional-
awareness within social work. The resulting collective-awareness potentially can go far
in enhancing professional social work education. Over twenty ﬁve years ago, Pilsecker
issued a call for the involvement of social work educators in such an undertaking. He
suggested that educators are “obligated to make it possible for the process to begin”
(1978, pg. 55). His call for educators to begin the process has been heeded through the
self-study process and institutionalized through the CSWE accreditation process.
However, the process remains unique to individual social work education programs. A
collective awareness is yet to develop throughout social work education.

A more recent call for educators to pursue evidence-based practice has been made
by Gambrill (1999). Social work educators indeed are in a unique position to initiate and
facilitate a self-study process, not only for social work education, but also for the

profession as a whole. To this end, social work educators need to frame professional

education in terms of evidence-based practice principles and craft an outcome evaluation
for demonstrating the profession’s achievement of its stated purpose and goals.

Recent changes by the Council on Social Work Education (CSWE) require social
work education programs to demonstrate compliance with educational policies,
adherence to education standards and achievement of educational objectives (CSWE,
2003). The educational policies and accreditation standards have resulted in routine
process evaluation and education quality improvement since the early 1990’s (CSWE,
1998). The 1994 evaluative standards, 1.4 and 1.5, required schools of social work to
engage in systematic self-assessment (CSWE, 1998). Current CSWE educational
policies and accreditation standards 8.0 and 8.1, enhance the earlier standards and call for
demonstrable outcomes in accredited professional social work education programs
(CSWE, 2003). This requirement has led to a ﬂurry of activity in undergraduate and
graduate schools of social work aimed at enumerating educational objectives, activities,
and outcomes (Holden, Meenaghan, Anastas, & Metrey, 2002; Miah & Newcomb, 1995).

Demonstrating outcomes necessarily requires enumerating attainable, relevant
goals, operationalizing objectives, and selecting valid outcome measures. The study
described in this dissertation was designed to explore one potential outcome measure for
evaluating social work education outcomes related to a core value of social work: the
profession’s commitment to promoting social justice and alleviating poverty. Work with
disadvantaged populations is an ethical mandate for social work practitioners and a social
justice curricular mandate for social work educators (NASW, 2000; CSWE, 1998). As
such, social work educators must demonstrate that social work graduates also share the
profession’s commitment to social justice and have the knowledge, skills, and values to

promote social justice and work to alleviate poverty.

Social work students’ attitudes about the poor and their beliefs about the causes of
poverty should be of critical concern to educators. As noted above, attitudes that
graduates bring to their work with clients affect individuals’ experiences with social
welfare. The relationship between attitudes and service delivery also affects public
perceptions of the social work profession, as well as, perceptions of poverty as a social
justice issue. Because of this complex of associations, social work educators must
evaluate the nature and strength of professionals’ knowledge, skills and values regarding
the issues, individuals, and communities that are targets of social work interventions.

As indicated in the statement of its purposes, “social work education is grounded
in the profession’s history, purposes, and philosophy” (CSWE, 2003, p. 6). When
evaluating professional education, it is important that we base identiﬁed outcomes on the
profession’s history, its stated purposes, and the philosophy threaded throughout both
history and purpose. Identifying social work’s philosophy, with regard to social injustice
and, in particular, poverty, is a difﬁcult task. The difﬁculty, in part, arises from a the
premise that the profession maintains a shared, unitary view of social injustices. A valid
contention, that the profession, in fact, does not have a unitary philosophy, provides the
ﬁrst conceptual challenge to the argument advanced in this dissertation. Additional ,
discussion of this study limitation is provided in the concluding chapter. However, the
magnitude of the difﬁculty does not relieve evaluators of the task of conceptually
deﬁning and operationalizing outcomes that allow for comprehensive assessment of
knowledge, values, and skills.

Professional Social Work Education
As speciﬁed by the CSWE (2003) the purpose of social work education is to

prepare competent professionals that have the knowledge, skills, and values to provide

effective social services in ways that reﬂect the purpose and philosophy of social work.

The National Association of Social Workers Code of Ethics states in its preamble:

 

The primary mission of the social work profession is to enhance human
well-being and help meet the basic human needs of all people, with
particular attention to the needs and empowerment of people who are
vulnerable, oppressed, and living in poverty. Fundamental to social
work is attention to the environmental forces that create, contribute to, and

address problems in living. [Social workers] strive to end
discrimination, oppression, poverty, and other forms of social injustice”
(1997,p.l)

Relatedly, the Council on Social Work Education stated within the Educational Policy
Statement for Bachelor of Social Work (BSW) and Master of Social Work (MSW)
programs

The purposes of social work education are to prepare competent and

effective professionals to develop social work knowledge, and to provide

leadership in the development of service delivery systems. Social work

education is grounded in the profession’s history, purposes, and philosophy

and is based on a body of knowledge, values, and skills. Social work

education enables students to integrate the knowledge, values, and skills of

the social work profession for competent practice (2003, p. 6).

These two passages provide a succinct picture of the goals and outcomes expected from
social work education. The need for meeting these goals is reﬂected in the post-
graduation market for trained social workers.

Every year, bachelor’s and master’s level social work graduates join the ranks of
professional social workers that provide government supported or contracted services to
disadvantaged families (NASW, 2005). One major employer of social workers is the
nation’s public welfare system, administered through the states. In addition to public
welfare programs, protective services workers for children and adults, foster care and

delinquency workers, preventive services workers and family preservationists provide

services to many individuals and families struggling with poverty and its correlates.

Given the breadth of involvement that social workers have with individuals,
families, communities, and other professions, it is crucial that we, as social work
professionals, understand our own attitudes toward impoverished clients. Attitudes that
graduates bring to their work with poor clients will affect the quality of their service
delivery and their clients’ experiences with the welfare system (Wyers, 1978; Piliavin,
Masters & Corbett, 1979; Mofﬁc, Brochstein, Blattstein & Adams, 1983; Reeser &
Epstein, 1987; Kingﬁsher, 1996; Rehner, Ishee, Salloum & Velasques, 1997). Therefore,
assessing social work students’ attitudes about the poor and their beliefs about the causes
of poverty and the nature of social justice must be of critical concern to educators.

Before we can evaluate student attitudes, however, we must begin with a review
of the profession’s history with respect to poverty and the poor. Social work educators
must assess student values and attitudes within the context of the profession’s history.
This review is especially important given social work’s ambivalence with regard to
poverty and the poor (Walkowitz, 1999; Jansson, 2001; Margolin, 1997; Trattner, 1994).
The following section presents a concise overview of our history with respect to poverty
and the impoverished.

Social work history, poverty, and the poor

In the United States the social work profession historically has maintained a
central position in societal care of the poor (Specht & Courtney, 1994; Jansson, 1988;
Roff, Adams & Klemmack, 1984; Morris, 1977; Margolin, 1997). Late 1700’s and early
1800’s anti-pauperism movements, and the growth of community administered
almshouses and poor farms, predated the Charitable Organization Societies (COS) and
settlement house movement activities later in the century (Day, 1997; Leiby, 1978;

Leiby, 1987; Trattner, 1984). The growth in “scientiﬁc charity” followed the established

traditions of private philanthropy and religious organizations as primary societal vehicles
for addressing pauperism (Day, 1997; Trattner, 1984). The poor originally were
recipients of social workers’ help through the charitable organization societies of the
1800’s. COS and settlement houses targeted the poor for services and intervention
(Addams, 1961; Jansson, 1988; Watson, 1922; Stuart, 1990).

The founders of professional social work furthered the young profession’s
involvement with poverty as a social justice issue. Despite their broad range of activities
and the variety of beliefs held by these early social workers, the overarching social
condition and population to which they attended was poverty and the poor. Mary
Richmond’s publication in 1899, Friendly visiting among the poor, predated the social
casework method she developed (Day, 1997; Leighninger, 1987). Richmond’s
conception of social casework grew from her own work with the poor and friendly
visitors that worked with impoverished families through charity organization societies
(Jansson, 1988; Longres, 1987).

Likewise, Lillian Wald’s work at Henry House (Edwards, 1987) and Jane
Addams’ work at Hull House (Quam, 1987), and their efforts on behalf of the settlement
house movement, were fueled by their individual commitments to changing social
conditions for the impoverished and exploited in urban centers throughout the United
States (Addams, 1961; Evans, 1997; Jansson, 1988; Stuart, 1990).

Edith Abbott’s belief in the govemment’s responsibility to address social
problems, furnish humane public welfare systems, and consider the social ramiﬁcations
of local and federal legislation further expanded the new profession’s concern with
poverty and the poor. The strenuous efforts of Bertha Capen Reynolds, on behalf of the

working poor and other oppressed groups, and the Rank and File Movement, delimited

more radical, grassroots, branches of the social work profession (Burghardt & Fabricant,
1987; Quam, 1987; Reynolds, 1963; Specht & Courtney, 1994).

The social work profession emerged during the 1800’s out of private
philanthropic work and religious efforts to alter the nature and extent of poverty
(Trattner, 1984; Leiby, 1987; Lunbeck, 1994; Walkowitz, 1999). From its genesis, the
profession identiﬁed the poor, worked with impoverished families and individuals, and
crafted policies to address poverty. Over time the profession acquired an integral
position in the creation and dispensing of private and public assistance (Walkowitz,
1999).

Social work’s history with regard to poverty and the poor, as described above, is a
patchwork of good intentions, misguided kindness, and “cruel compassion” (Margolin,
1997; Szasz, 1994). Margolin draws upon a 100-year period of social work literature to
illustrate the semantic ways in which the profession has “blamed the poor for their own
poverty” (p. 97). The literature he examines contains both 19th and 20th century treatises
on the characteristics of the poor and justiﬁcations for intrusive forms of public
assistance. Margolin (1997) cites a passage from 1964, in which Levine describes
attributes of the poor:

Conceptual or abstract thinking conveyed through words and phrases is

beyond the capacities of this unsophisticated group, members of which are

marginally educated, whose lives are a social and cultural wasteland, who

act out anger and hostility, who have a low frustration tolerance and poor

impulse control. Moreover, in the formative years of the parents and now

their children, words were and are used to manipulate and confuse others.

These people do not comprehend the true meaning of words, have little faith

in them, and are unable to carry out concepts deﬁned by words alone (p. 97)

Levine’s work, entitled Treatment in the home, was published in the journal Social Work.

Passages like the above are not anomalies in the professional literature, as Margolin

(1999) notes, but rather, reﬂect a widespread sentiment within the profession. The
sentiment has been, and can be still, traced throughout the profession’s history (Gans,
1995; Lunbeck, 1994; Margolin, 1997; Specht and Courtney, 1994; Walkowitz, 1999).

Since the beginning, professional social workers have maintained a variety of
opinions or beliefs about poverty and the poor. Some have viewed poverty as a
consequence of capitalist industrialization and the exploitation of children and
immigrants. Others have considered poverty to be the result of rampant alcoholism and
the abandonment by husbands and fathers of their wives and children. Despite the
diversity of beliefs, within the COS there was broad-based agreement that the underlying
issue in poverty concerned morality and civility. Moral turpitude, laziness, spendthrift
habits and a poverty of culture were addressed with punitive language and moralistic
judgments (Jansson, 1988; Lunbeck, 1994; Margolin, 1997). The presumed dishonesty,
avarice, and ignorance of impoverished families and individuals was a driving force
behind the “scientiﬁc charity” of COS. The COS coordinated their records to ensure that
recipients of their philanthropy were not beneﬁting ﬁ'om duplicative charity (Jansson,
2001; Margolin, 1997).

The settlement houses throughout the United States held a divergent view of the
causes and issues of poverty. However, the sentiment and methods behind the settlement
houses included efforts to provide the urban poor with more than just basic needs.
Cultural enrichment and access to middle-class civility were integral parts of settlement
house activities. Although social workers’ were concerned with improving conditions
and enhancing the lives of the poor, their underlying beliefs were based on attributions
about individual characteristics of the poor that sustained their poverty or individual

experiences that might ameliorate its effects (J ansson, 2001; Lunbeck, 1994; Walkowitz,

1999). Over time, the settlement house movement transformed into a professionalized,
community-based arena for group work, referral, and recreation-oriented programming
(Stuart, 1990).

The historic relevance of beliefs about poverty and attitudes toward the poor for
the profession of social work cannot be overstated. The profession’s history is peppered
with good intentions closely followed by moralistic judgments and unsubstantiated
attributions (Margolin, 1997; Piven & Cloward; 1993; Trattner, 1994; Walkowitz, 1999).
Our history provides both a mandate and direction for contemporary social work
education. Social work educators need to articulate the espoused values of the profession
over time and systematically examine held values and attitudes of the new professionals
we educate. The next section explores in more detail the means by which educators can
evaluate the achievement of educational outcomes with respect to poverty and the poor.

Evaluating social work education outcomes

The means for self-study already are present in social work education. The
accreditation standards and educational policies ﬁrmished by the CSWE (2003) require
periodic review of social work education programs. This requirement has created a
regular process in every accredited professional program whereby educators examine
their curricula to demonstrate achievement of the standards and compliance with the
stated policies. The accreditation process has evolved over time to its current state, one
in which professional programs must demonstrate having attained their stated outcomes.
Requiring empirical evaluation of education outcomes, in additional to demonstrating
compliance with policies, presents new challenges to social work educators.

The requirement for outcome evaluation is equivalent to the challenges made to

social work practitioners by scholars during the last twenty ﬁve years, namely, to engage

10

less in authority-based practice and more in evidence-based practice (Gambrill, 1999;
Gambrill, 2001). Since the 1980’s social work scholars have been calling for
professional education and service delivery to embrace evidence-based practice (Fischer,
1983; Gambrill, 1999, 2001; Cournoyer, 2004).
Evidence-based Practice

By conceiving of social work educators as practitioners working in an academic
setting, the strictures of evidence-based practice can be applied to the problem of
evaluating professional education outcomes. Social work education can be viewed as
another ﬁeld of practice, similar in some respects to child welfare, gerontological social
work, or mental health services. The value in redeﬁning professional education as a ﬁeld
of practice comes from the application of principles of evidence—based practice.
Coumoyer (2004) recently proposed a deﬁnition of evidence-based social work:

Evidence-based social work is the mindful and systematic identiﬁcation,

analysis, evaluation, and synthesis of evidence of practice effectiveness as

a primary part of an integrative and collaborative process concerning the

selection and application of service to members of target client groups.

The evidence-based decision-making process includes consideration of

professional ethics and experience as well as the personal and cultural

values and judgments of consumers. (p. 4)
This deﬁnition includes several ideas, which require elaboration with respect to social
work education as an arena of social work practice, the concept of effectiveness, the
nature of evidence, and the speciﬁcation of the targeted client group.

Effective Social Work Education

Speciﬁcally, the effectiveness of professional education can be described as “the

extent to which services yield their intended results” (Cournoyer, 2004, p. 4). This

deﬁnition requires social work educators to operationalize the “intended results” of

professional education. The question becomes, “What should we teach/know about

11

poverty?” In order to understand the outcomes of social work education it is necessary to
turn to the Council on Social Work Education (CSWE). In the 2003 Educational Policy
and Accreditation Standards (EPAS) the CSWE states:
The purposes of social work education are to prepare competent and
effective professionals to develop social work knowledge, and to provide
leadership in the development of service delivery systems. Social work
education is grounded in the profession’s history, purposes, and
philosophy and is based on a body of knowledge, values, and skills.

Social work education enables students to integrate the knowledge, values,
and skills of the social work profession for competent practice (2003, p.6).

As enumerated in the Educational Policy and Accreditation Standards (EPAS), the
profession’s purposes include two statements of particular relevance to this study:
enhancing human well-being and alleviating poverty, oppression, and other forms of
social injustice; pursuing policies, services, and resources through advocacy and social or
political actions that promote social and economic justice. Fulﬁlling the profession’s
purposes, particularly with respect to poverty and economic justice, ultimately will reﬂect
the effectiveness of social work education and professional practice.
The nature of evidence of educational effectiveness

As educators we are familiar with evaluating student acquisition of knowledge,
typically achieved through testing or grading written assignments that apply newly
acquired knowledge. Within social work, educators are familiar with teaching and
assessing skill development among students undergoing ﬁeld education. As a profession
we are in relative agreement about the breadth of knowledge and level of skill a newly
minted social worker should have. And, we have devised ways to evaluate both
knowledge and skill throughout a student’s education. We have had less success in
routinely assessing values students hold when they enter professional education and as

they graduate.

12

Transformation or enhancement of student values, which results from social work
education, largely has been left to faith. That change happens, and that the resulting
values, beliefs, and attitudes are consonant with the profession’s values, has been more
assumption than veriﬁable fact. Assessing held values, and evaluating changes in values,
is important because, as Piven and Cloward (1993) have noted, “professional beliefs
serve the same economic and political functions as philanthropic beliefs in guarding the
system against claimants for aid” (p. 175). The need for actively deﬁning and
operationalizing the profession’s values to a degree that renders them measurable is
profound and, as yet, unmet.

The profession already has delineated its values. The NASW Code of Ethics

(2000) enumerates six core values:

Service

Social justice

Dignity and worth of the person
Importance of human relationships
Integrity

Competence

The Code of Ethics states, “These core values, embraced by social workers throughout
the profession’s history, are the foundation of social work’s unique purpose and
perspective (p. 1). The challenge becomes how to measure values. However, one
additional aspect of evidence-based practice remains to be delineated.

The Code of Ethics provides not only an account of core social work values, it
also furnishes broad ethical principles based on the values, and sets ethical standards for
professional conduct and resolution of ethical dilemmas. The function of the code is to

describe and support an overall pattern of professional behavior, both for individual

13

social workers and for the profession as a whole. Among the stated purposes of the
NASW (2000) code the following are described:

0 The Code provides ethical standards to which the general public can hold
the social work profession accountable;

- The Code socializes practitioners new to the ﬁeld to social work’s
mission, values, ethical principles, and ethical standards (p. 2).

These stated purposes add greater urgency for meeting the need for evaluation of
social work education. The purpose statements also reinforce the importance of
assessing values and value changes with respect to professional education.
Target population

Evidence-based practice necessarily entails identiﬁcation of whose practice for
whose beneﬁt. The consumers of the social work profession’s services generally are
thought to be the individuals, families, organizations, and communities on whose behalf
professionals work to enhance well-being and alleviate social injustice. However,
applying an evidence-based practice model to social work education, reveals that social
work educators, and the educational process, are working to enhance the well-being of
the social work profession. Social work educators serve a population comprised of
existing and future professionals and organizations, professionals who, in turn, serve the
profession’s customary client groups, individuals, families and communities.

Evidence-based Education

Despite NASW’s broad enumeration of social work core values, as educators and
evaluators we are left the task of operationalizing those values and choosing or crafting
valid measures. Evaluation research has grown into a sophisticated methodology for
building the social service knowledge base (IASWR, 2001; Rossi, Freeman and Lipsey,

1999). Recent developments have yielded a useful evaluation tool, the Logic Model

14

(Mulroy and Lauber, 2004). Logic models conceptually facilitate organization of
evaluations of health, education, and social welfare services (Kapp, 2004). Logic models
offer a method for structuring an evaluation of social work education, in general, and of
social work students’ attitudes, in particular. The following section details the six
components of a logic model structure proposed by Kapp (2004). The model is described
with respect to graduate education in schools of social work.
Logic model for social work education

A logic model for an outcome evaluation of a program can be conceived as
having six components. Figure 1 details the structure and content of a logic model for
use in evaluating social work education. Logic modeling describes three aspects of any
program, inputs, outputs, and outcomes. The ﬁrst three components of the proposed
logic model constitute the inputs to social work education, (1) resources, (2) faculty
activities, and (3) program processes. The next two model components constitute the
outputs generated by social work education, (4) immediate outcomes and (5) intermediate
outcomes. The ﬁnal model component is (6) long-range outcomes.
Inputs

Resources. The ﬁrst model component enumerates resources. Program resources
in social work education are the concrete inputs, i.e., students, faculty, a school’s physical
environment or facilities, courses, textbooks and other scholarly works, non-classroom
based events, and tuition.

Faculty activities. The second, major input involves faculty activities in support
of the school’s mission. Faculty activities in professional social work education include
teaching, research, social work practice, committee work, school governance meetings,

and community involvement. Teaching is a multifaceted technique for educating

15

professionals. Not only does teaching allow educators to introduce new knowledge to
students, teaching also provides an opportunity for educators to model social work values
and ethical principles. Students also learn through critical review of their thinking and
writing elicited by course assignments. Challenges to expressions and interpretations
inconsistent with social work values can be made during classroom contact or through
feedback on assignments or group activities. Reinforcement of professional thought and
behavior also is possible in classroom settings.

Program processes. The third input category, program processes, is comprised of
the procedures by which social work schools operate and through which students are
transformed into professional social workers. Teaching is a primary process in
professional education. Educators work within a classroom setting where they teach to
approved course objectives listed in course syllabi. The objectives reﬂect the educational
policies and accreditation standards issued by the Council on Social Work Education. In
addition to teaching, and student evaluation through grading, schools of social work
structure their curricula in uniform ways. Every professional school includes a
curriculum with demonstrable content on values and ethics, diversity, 'at—risk populations
and social and economic justice, human behavior in the social environment, social
welfare policy and services, social work practice, research, and ﬁeld education. Schools
also offer advanced content with “greater depth, breadth, and speciﬁcity”, with respect to
ﬁelds of practice, than is covered in the foundation content (CSWE, 2003, p. 12).

These three inputs for social work education, resources, faculty activity, and
program processes, typically are assessed through established plans and procedures for

ongoing reviews within each social work program. Additionally, professional schools are

16

$88 0&2 8:835 x83 Eoom ._ oSmE

 

powwow 38m
:oEEo 333m
3255 Etna
28m 58?.

88830
owned weed

 

mmEOUPDO

23:3: .moﬁsvﬁw
muoumozmmm 68on 8982386 :ouﬁmuouooxx
mos—g Eoc< 323266 3800
25m mo_o>oQ Seton—go 22m

ammo—305— 2333. 338023 wcgﬂ 20E

 

“~033me $50 :8me
88880
BEBEBE 880830 8368c:
mHDAEbO

5:833 mewsﬁﬁou

:ouSmcouoomém 853.556
SERVED mooEEEoO
=28:me 23m 838$
3680 subsumed
wanmoH wanmoH

 

 

8&8on Badman 8:333. 330%

”:57:

$55 Begum
mEpEoomE 20E
3on

momBoU
Sam—moan
35an

page

moousomom

 

17

evaluated on a periodic basis through the CSWE reaccreditations process. These inputs
are intended to yield the outcomes detailed above in Figure 1.
Outputs

Immediate outcomes. Social work education outputs are two-fold, immediate and
intermediate. Immediate outcomes are straightforward, customary indicators of
educational attainment. Student grades on assignments and for completed courses are
primary measures of educational effectiveness with respect to knowledge. In addition to
grades, the content and achievement of ﬁeld learning agreements or contracts furnish
qualitative measures of student skills, knowledge application, and ethical behavior.
Despite this breadth of coverage there is no systematic, direct measure, either qualitative
or quantitative, of student values upon entering graduate social work programs or value
changes throughout graduate education.

Intermediate outcomes. A variety of gross measures of educational policy
compliance and achievement of accreditation standards comprise intermediate outcomes.
Periodic reaccreditation is a program-based assessment of compliance and standards
achievement. Graduation of MSW credentialed social workers that are able to pass
licensure or certiﬁcation requirements provides another intermediate measure of
education success. Increased applications for study and growth in student enrollment
also serve as measures of educational effectiveness.

Students’ knowledge, skills, and values are other measures of intermediate
outcomes. In some schools of social work, student course requirements include an
integrative seminar at the end of graduate study. Such requirements allow schools to

ascertain student achievement of educational objectives beyond assessments of course

18

objectives — typically measured by individual course grades and overall grade point
averages.
Outcomes

Long range outcomes. Long-range outcomes reﬂect the purpose of social work
education and the mission of the social work profession. If social work education is
effective, schools should be able to demonstrate they have created social workers that are
prepared, competent professionals who develop social work knowledge, provide
leadership in service development and delivery, and who integrate the knowledge, values
and skills of the profession in their practice (CSWE, 2003). Within any given school of
social work these professionals constitute the alumni.

Meeting the mission of the social work profession also is a longer-term outcome.
The mission includes enhanced social and economic justice, decreased social injustice,
poverty, and oppression; increases in meeting the needs of individuals and groups; and
advocacy and political actions to expand policies, services, and resources with respect to
individual well being, poverty, and social justice (NASW, 2000). The profession’s
activities in pursuit of its mission also should be reﬂected in a more positive public
opinion of social workers and the social work profession as a whole.

Overview of Logic Model for Evidence-Based Education

The logic model presented above provides an overview of social work education
inputs and outcomes. A detailed evaluation plan would be used to address sequentially
each outcome, identify contributing activities, and suggest measures or indicators of

inputs and outcomes. This task is beyond the scope of the study. However, one aspect of

19

the logic model can be developed as a circumscribed example, and an initial contribution,
to the evaluation of social work education as an evidence-based practice.
Student values as an outcome of eﬂective educational practice

Adoption of social work values is among the most challenging educational
outcomes to measure. Student knowledge acquisition and professionals’ knowledge
development over time can be assessed directly. Testing for licensure or certiﬁcation,
now used in 35 states, is a means for assessing professional knowledge (Bowden, 2005).
Social worker skills are concrete manifestations of the knowledge and experiences of
novice and master social workers. Novice social workers’ skill levels are monitored
through ﬁeld instruction and close supervision by seasoned social workers and
administrators. Master social workers typically practice in organizations with program
and agency standards, routine supervising, and accountability procedures.

Unlike knowledge and skill, the acquisition, adoption, or reﬁnement of values is
far more difﬁcult to operationalize and measure. Assessing the application of those
values also poses a challenge because it requires speciﬁcation of a target about which
values are concerned or to which values are applied. This study focused on poverty and
the poor as the referent for values because of the prominence of poverty, both historically
and currently, as a concern for the social work profession. This dissertation study assists
with the task of operationalizing social work values by reﬁning a relevant measure.

The Attitudes Toward Poverty (ATP) scale, proposed by Atherton, Gemmel,
Haagenstad, Holt, Jensen, O’Haran, and Rehner (1993), was developed to be a credible
tool for measuring attitudes and deducing underlying values held by social work students

and professionals. The next chapter presents a brief argument to validate the

20

measurement of values using attitudes as values indicators. This rationale for the study is
followed by a review of the empirical literature on social work student attitudes about
poverty. Based on a review of empirical literature, possible correlates of social work
student attitudes about poverty and the poor were identiﬁed. The resulting list of
variables provided direction for selecting demographic and personal background
questions to include in the survey. Deﬁciencies in existing research are enumerated.
Research questions, following from variables identiﬁed as possible correlates, are
enumerated.

Chapter 3 details the methodology and analysis approach used to examine the
ATP and explore the research questions. In Chapter 4, the presentation of ﬁndings begins
with the replication of the original ATP scale. Scale reliability analysis is followed by
conceptual and empirical justiﬁcations for reﬁnements to the scale. The enhanced ATP
instrument is used to address the research questions posed at the end of the review of the
literature. Chapter 5 discusses the ﬁndings with respect to the research questions.
Although the study was not designed to test theory, theoretical frameworks are drawn
upon to facilitate interpretation of ﬁndings. The dissertation concludes with discussion of
the implications of the study for evidence-based education, limitations of the study, and

suggestions for further research.

21

CHAPTER 2
REVIEW OF THE LITERATURE

This chapter reviews the literature related to the measurement of social work
student attitudes toward poverty and the poor. The chapter begins with a brief review of
the theoretical context of attitude measurement, attitude formation, and the connection
between values and attitudes. This contextual overview is followed by a review of
empirical literature on social work student attitudes toward poverty and the poor. This
examination of knowledge regarding student attitudes about poverty is followed by a
brief critique of the research. The chapter concludes with a presentation of the research
questions examined in the current study.

Knowledge, Values, Attitudes and Behavior

The study of attitudes in social psychology and sociology suggests a means for
creating meaningful evidence for evaluating social work education. Knowledge and
values are reﬂected in attitudes (Cournoyer, 1991; F ishbein & Azjen, 1975). And, in
turn, attitudes affect behavior (Carrillo, Holzhalb, & Thyer, 1993). Azjen and F ishbein
(1980) and Fishbein and Azj en (1975) provided comprehensive reviews of attitude
research. Social science research on attitudes has a long history, predating the formation
of the social work profession (Azjen & Fishbein, 1980). Contemporary developments in
social science knowledge about attitudes reﬂect “agreement among investigators that
attitudes toward any object are determined by beliefs about that object” (1980, p. 62).
Azjen and Fishbein maintain, “in order to understand why a person holds a certain
attitude toward an object it is necessary to assess his salient beliefs about that object”

(1980, p. 63). They also observed, “quite simply, an attitude is an index of the degree to

22

which a person like or dislikes” the focus of the attitude. In the current study the
attitudinal object was poverty and the poor.

Azjen and Fishbein (1977) noted in their review of research on attitudes and
behavior that “attitudes toward an object can predict only the overall pattern of behavior”
(p. 27). They cited research by Thurstone (1931), Doob (1947), and Campbell (1963), in
support of their contention that speciﬁc behavior toward an object cannot be predicted by
knowing only an individual’s attitude toward that object. This observation suggests that
in the current study we may not conclude that negative or positive attitudes toward
poverty and the poor will result in particular treatment of impoverished individuals,
families or communities. However, it lends credence to the use of attitude measurement
as a means for tapping underlying beliefs and values that may inﬂuence an “overall
pattern of behavior” in students and, eventually, credentialed professionals.

Rokeach (1968) deﬁnes attitude as, “a relatively enduring organization of beliefs
around an object or situation predisposing one to respond in some preferential manner”
(p. 112). Attitudes differ from beliefs in that “beliefs relate less to one’s disposition or
learning than to one’s assessment of an empirical situation” (Tropman, 1989, p. xiv).
Beliefs relate more closely to knowledge, facts, and evidence. Attitudes, however, can be
maintained in the absence of empirical support. If attitudes are thought of as proxy
indicators of individuals’ beliefs and values, then assessing student attitudes should
provide an acceptable measure of educational effectiveness for increasing knowledge and
inﬂuencing social work student values in the desired direction.

More recently poverty attitude research with social work students reviewed

theories of attitude formation (Melnick, 2000). Deﬁnitions of attitudes, with importance

23

for the current study, are credited to three researchers, Rokeach, Sherrod, and Gray.
Rokeach, in 1968, articulated three assumptions about beliefs for any given individual:

“Not all beliefs are equally important”

“The more central a belief, the more it will resist change”

“The more central the belief changed, the more widespread the

repercussions in the rest of the belief system. (p. 3)” (as cited in
Melnick, 2000, p. 15)
Melnick (2000) noted, “attitudes are important as they relate to a person’s behavior.
While attitudes themselves do not constitute behaviors, they are ‘mental processes that
related individuals to objects’ (Gray, 1996, paragraph 4)” (Melnick, 2000, p. 19).
Melnick also cites “‘Sherrod (1982) [who] notes that it is easier to predict multiple
behaviors from a general attitude than to predict a single behavior’” (2000, p. 19).
Research On Social Work Student Attitudes Toward Poverty And The Poor
A number of studies have been published since the early 1960’s regarding student

attitudes toward the poor or beliefs about the causes or nature of poverty. A 1974 study
by Sharwell, reportedly used a sample of 20 students. During the course of their social
work education, students experienced signiﬁcant increases in their initially positive
attitudes toward public dependency. Orten (1979) found an extensive literature on
attitudes toward the poor. Of the studies he reviewed (Cryns, 1977; Grimm & Orten,
1973; Orten, 1976), social work students’ demographic characteristics were differentially
related to their generally positive attitudes toward poverty and the poor. Orten’s review
(1979) of studies by Varley (1963) and Hayes and Varley (1965) noted no changes in

student attitudes during graduate social work training. However, the authors had

indicated that students who entered graduate school with practice experience exhibited an

24

attitude shift by the end of their professional training such that they viewed the poor more
negatively and attributed poverty to personal failing rather than social injustices.

Orten (1979, 1981) reported that social, economic and demographic
characteristics did not account for actual attitude but were somewhat related to the
intensity with which attitudes were held. Orten indicated that change in attitude,
following students’ participation in a poverty simulation, did not occur for Black students
or for students from lower socioeconomic levels.

Several other studies (Hayes & Varley, 1965; Varley, 1963) were designed to assess
changes in student attitudes toward the poor during social work training. In one study,
Orten (1976) reported a subgroup of students - those with initially more negative attitudes
- became more negative by the end of their social work education; for the group overall,
attitudes became more positive.

Throughout the 1980’s, another four studies on social work student attitudes
toward poverty and the poor were published. Macarov (1981) reported on a multi-
national study of social work students’ attitudes about poverty deﬁnitions, causes, and
remedies. BSW and MSW students in Australia, Israel and the United States are
described as having “confused, conﬂicting and ambiguous” attitudes toward poverty and
the poor. Australian students tended toward a “conceptual” deﬁnition and understanding
of poverty and endorsed system changes to alleviate poverty. Israeli students tended to
attribute poverty more to personal “deﬁciencies” and recommended education as a
curative measure. United States students were more “practical” when deﬁning and

attributing causes to poverty; they offered employment as a solution. Macarov

25

recommends that students be given more facts about poverty and policies for alleviating
it and be assisted in identifying and modifying their own attitudes.

Mofﬁc, Brochstein, Blattstein, and Adams (1983) advanced the position that
physical and mental health providers in the public sector need sensitivity to the needs of
poor patients. To this end, the researchers crafted an “innovative inter-institutional, inter-
agency, and interdisciplinary primary care /community mental health training and service
program” (p. 19). The program included 18 ﬁrst and second year graduate social work
students and 24 psychiatric resident physicians. Both medical and social work students
were assigned to one of ﬁve public mental health clinics that were neighborhood-based
and provided mental health care to underprivileged patients. The study assessed trainees’
attitudes toward “patients of lower social class” using Lemer’s (1973) Democratic Values
Scale (p. 19). The Democratic Values Scale (DVS) provided a measure of “the depth and
genuineness of a therapist’s commitment to democratic values” (p. 19), which assessed
the willingness of a respondent to use coercion with clients and willingness to limit client
autonomy. The authors note the construct validity of the DVS as reported by Lerner
(1973). Findings indicated the training model was effective in positively increasing
democratic values from pretest to posttest. Despite the low sample sizes, and relatively
small numbers of minority trainees, race status and change from pre-test to post-test
showed an interaction effect. Mofﬁc, Brochstein, Blattstein, and Adams concluded “that
social work trainees may develop attitudes particularly conducive to working
therapeutically with poor patients” (p. 24). They recommend that future research assess
demographic factors such as “ethnicity, sex, age, religious preference, marital status, and

personality” (p. 27).

26

Roff, Adams and Klemmack (1984) reported their comparison of MSW students
and students enrolled in Introduction to Sociology classes. They found support for the
hypothesis that MSW students were more supportive of the idea of government helping
the poor when a reason for their poverty was not considered. Their willingness to have
government help the poor dropped signiﬁcantly when students could attribute poverty to
an internal or intentional reason on the part of the poor person. Finally, MSW students
agreed more with statements about poverty being caused by economics and
discrimination and less with the statements that the poor are not willing to work hard.
There were no differences between social work and sociology students for the other
poverty explanations advanced by the researchers - the poor turn down jobs, lack job
skills, have health problems, are laid off, have outdated job skills. Roff, Adams and
Klemmack concluded that MSW students hold beliefs about deserving and undeserving
poor. They suggest that social work students should be explicitly taught about the many
causes of poverty and about the social work value of “access to adequate resources
regardless of the worker’s personal feelings” (1984, p.19).

Related to research on student attitudes, a study by Reeser and Epstein (1987)
examined changes in surveys of social work professionals’ attitudes toward poverty and
social action between 1968 and 1984. Reeser and Epstein posed the research questions,
“Where is social work today in regard to the poor and social action?” and have social
worker attitudes about the poor and social action changed since the 1960’s. Their work
ﬁlled a void in scientiﬁcally sound, empirical research regarding social worker attitude
changes over two decades. The study compared survey data from a 1960’s era study by

Epstein (1968) with survey data from a replication of Epstein’s study conducted by

27

Reeser (1984). Epstein’s 1968 study included a probability sample of members of the
New York City chapter of NASW; sampling yielded an approximate sample size of 670.

Reeser used probability-based sampling from the national NASW membership
list. Reeser’s approximate sample size was 760 respondents. Both the 1968 and 1984
samples were statistically representative of the larger sampling frame. Social work
students were omitted from the 1984 study. The samples differed on a variety of social
dimensions: agency position, ﬁeld of service, agency auspice, race, and religion; there
was no observed difference for gender. Attitudes toward poverty were measured by
eliciting from respondents the “two most important reasons for the existence of poverty
in the US.” (p. 613). Answer choices were as follows:

0 Individualistic -— “Poor people are not adequately motivated to take
advantage of existing opportunities”

0 Social structural — “Powerful interests are fundamentally opposed

to the solution of the problem of poverty” and “Those people who
are better off will never give up anything unless forced”

o Technological — “We do not possess the necessary knowledge and
techniques”

0 Interest group —- “People representing different interests do not
often enough sit down together to work out the problem” and
“Poor people have not been organized to demand better treatment
by society”.
Two additional items were rated for agreement:
0 “The only way to do away with poverty is to make basic changes in our
political and economic system”.
0 “The poor are in the best position to decide what services they need”.
Signiﬁcant differences between the two samples were found for all

explanations of poverty and for both agreement items. The 1984 sample had

more frequent identiﬁcation of the social structural explanations. The 1968

28

sample more frequently identiﬁed the individualistic, technological, and interest
group reasons as explanations for poverty. The 1984 sample had a signiﬁcantly
higher proportion of agreement for both rated items, such that the later sample had
twenty percent more agreement (81%) than the earlier sample (61%) that basic
changes were needed in political and economic systems, and the later sample had
ﬁfteen percent more agreement (51%) than the earlier sample (35%) with the
belief that the poor know best what services they need. Respondents from 1968
preferred working predominantly with the poor more frequently than reported
preferences of the 1984 sample (14%).

Reeser and Epstein concluded that, in the 1980’s, social workers [did] not
“regard the elimination of poverty as a priority of the profession. Instead, they
view[ed] the role of social work as helping individuals of all social classes to
adapt to the environment” (p. 621). They suggested that social workers in the
1980’s “view[ed] social work as a ‘consenting’ profession supportive of the social
class system” (p. 621).

The volume of research on social work student attitudes increased sharply in the
1990’s. A combination of thesis and dissertation research, and peer reviewed studies
were published, spurred by growing concerns that social work had abandoned it’s
mission (Specht and Courtney, 1994). A number of these more recent studies examined
the effectiveness of various teaching methods, academic content, or learning activities on
altering student attitudes and beliefs about poverty and the poor (Santangelo, 1992;
Baggett, 1993; Beigi, 1998; Melnick, 2000). These studies are reviewed chronologically

in the remainder of this section.

29

Schwartz and Robinson (1991) reported on their study of beginning, intermediate
and advanced BSW students during the 1986-1987 academic year. Students were asked
to complete a poverty explanation survey and the 1966 Rotter Locus of Control Scale.
Students in all three levels of BSW education gave relative ratings to structural
explanations, i.e. discrimination, restricted opportunity, were rated the highest, followed
by fatalistic explanations, e.g. bad luck, and lastly, individualistic explanations, i.e.
character ﬂaws. Advanced students did not differ from either intermediate or beginning
BSW students on personal locus of control. Beginning students had a more internal,
personal locus of control than intermediate level students. The authors conclude that
these BSW students held attitudes toward the causes of poverty that are “congruent with
desired professional values” (1991, p. 294). They call for longitudinal study on the
impact of students’ attitudes on their professional activities. Continued research on the
impact of social work education on values about poverty and the poor is recommended.

Guttman and Cohen (1992) describe their investigation of Israeli students’
knowledge of, attitudes about causes of, experience with and willingness to work with
poverty. Undergraduate social work students were compared to students in economics,
sociology, political science and literature on the above variables. Undergraduate Israeli
social work students did not differ from the other student groups on any of the poverty
related variables. Guttman and Cohen conclude that in Israel, social work education has
“failed to produce the distinctive set of attitudes and values with regard to the poor on
which the profession has traditionally built its contribution to social progress” (p. 61).

Rosenthal’s (1993) research tested a growing belief within the profession that

social workers and social work students have abandoned the disadvantaged in

30

professional activity and attitude. In addition, she tested the hypothesis that social work
students have unrealistic beliefs about the actual ﬁnancial condition of being poor. She
developed three new measures which assess (1) respondent belief in individual causes
versus restricted opportunity explanations of poverty, (2) self-reported dislike for and
avoidance of the poor, and (3) estimates of yearly income for a variety of disadvantaged
recipients of government support. All study participants were MSW students. The
sample largely rejected individual causes of poverty and overwhelmingly reported a
willingness to associate with the poor. Student estimates of income were consistently
inaccurate and often much higher than aid recipients actually receive. None of the study
variables were signiﬁcantly related to background or demographic characteristics of
study participants. Rosenthal interprets her ﬁndings as being inconsistent with the
observation made by Specht and Courtney (1994) and others (e. g. Dean, 1977) that the
social work profession is dishonoring its legacy and abandoning its mandate to serve the
disadvantaged.

Atherton, Gemmel, Haagenstad, Holt, Jensen, O’Haran, and Rehner (1993)
published a brief report on their development of a scale for measuring social work
students’ attitudes toward poverty and the poor. These investigators offered the scale as a
“dependable” tool for social work researchers, educators, and public presenters to use “to
identify pertinent issues” and address “value questions associated with poverty” (p. 28).
The survey was tested and reﬁned using a convenience sample of 99 undergraduate social
work and sociology students and 113 undergraduate business students from three
universities. The sociology and social work students had signiﬁcantly more positive

scale scores than did the business students. Internal reliability and face validity of the

31

scale was acceptable, with Cronbachs’ alpha coefﬁcient of .83 and a Guttman split-half
coefﬁcient of .87. The study did not examine demographic variables such as age, gender,
education, income, ethnicity, religion, or political orientation. The authors called for
other educators, researchers and practitioners to use the scale and share results.

Beigi (1998) reported on research comparing social work students to accounting
students on attitudes and knowledge about public assistance and the poor. The sample
included 70 students from a large public university in California. Non-probability,
convenience sampling resulted in 37 social work student respondents and 31 accounting
student respondents. The survey consisted of demographic questions about age, gender,
marital status, ethnicity, annual household income, number of children, educational major
and standing, personal receipt of public assistance, and employment status. The survey
also contained 54 statements to which respondents indicated their agreement using a four-
point Likert type scale from strongly agree to strongly disagree. The survey contained
two categories of statements. The ﬁrst category concerned knowledge about three
aspects of public assistance, policies, environmental factors affecting public assistance,
and general knowledge about public assistance. The second categoryincluded items that
assessed attitudes related to stigma around public assistance and judgmental attitudes
toward the recipients of public assistance. The scales were not internally reliable and
items were analyzed individually. Overall social work students were less judgmental and
disagreed more strongly (M=3.18) that people on welfare are lazy when compared to
accounting students (M=2.80). And, social work students (M=2.89) disagreed
signiﬁcantly more than accounting students (M=2.21) that welfare should be given to all

people. Beigi concluded that social work and accounting students were comparable in

32

their knowledge and attitudes. Social work students did not have substantially more
knowledge or better attitudes. Beigi calls on social work educators for a curriculum to
change attitudes among social work students, and others.

Melnick (2000) reports research on the impact of social work curriculum on
student perceptions of the causes of poverty and the formation of attitudes about the poor.
Melnick used a quasi-experimental design, with pre-test and post-test to determine
whether graduate social work education affected graduate social work students’ attitudes
toward the poor. A sample of 52 respondents was composed of social work students
drawn from three schools of social work and a comparison group of graduate students in
arts, music, and education. The response rate for the mailed pre-test and post-test surveys
was low (47%) but not uncommon for mailed questionnaires. Results conﬁrmed the
hypothesis that professional socialization inﬂuences positive attitude change over the
course of graduate social work education. No changes in attitudes or beliefs about the
causes of poverty were noted for the comparison group. Socio-demographic variables,
related to the positive shift in attitudes in MSW students, included: pre-MSW social work
experience, age, gender, sexual preference, number of children, residence characteristics,
political party afﬁliation, and marital status. In addition, MSW students attributed the
positive changes in their attitudes toward the poor to faculty, ﬁeld education, and
coursework. Other demographic variables were unrelated to attitude changes.

Overview and critique of the research on students’ poverty attitudes

The reviewed studies provide a picture of social work students’ and professionals’

attitudes about various aspects of poverty as an issue of concern to the profession.

Observations about the existing research, offered by Santangelo (1992) remain relevant

33

and descriptive of the research to date. “Data gathered over the past three decades on
graduate social work students’ attitudes, values, and career aspirations have been
inconsistent, reﬂecting methodological variations and diverse areas of focus”
(Santangelo, abstract). The effort to address social work student beliefs, knowledge, and
practice interests has been ongoing since the 1960’s. The lack of uniformity in attitude
assessment tools, and the range in research methods, has led to a variety of gaps in our
knowledge about social work student attitudes toward poverty and the poor. The social
work profession has a distinct, unmet need for a poverty attitudes measurement tool
(Orten 1979). Such an instrument would enhance knowledge building and could
stimulate curricular innovations in graduate social work education.

Substantial gaps exist in the literature on student attitudes about the poor and the
causes and nature of poverty. First, a number of the studies contrasted social work
students with undergraduate and graduate students in other disciplines or professional
schools. Social work students consistently emerged as having more positive, less
pejorative attitudes and values with respect to poverty, the poor, and related policies and
programs. However, social work student attitudes were not compared routinely with an
attitudinal standard or position, or compared against social work faculty attitudes.
Although social work students maintain better attitudes toward economic disadvantage
and impoverished people than students from other ﬁelds of study, their attitudes may
remain less positive than educators or poor clients are satisﬁed with when interacting
with individual students.

Two additional gaps are evident; the second concerns an absence of studies that

examine potential associations between social work student attitudes and client

34

experiences or outcomes. If attitudes are to be a relevant construct about which educators
must be concerned about, then evidence should exist which demonstrates their
importance for client outcomes and service quality. Third, several studies (Melnick,
2000; Guttman & Cohen, 1992) tested the effects of professional education, speciﬁc
curricula, or other teaching innovations, with respect to knowledge or beliefs about
poverty and the poor. However, there was no uniformity in the attitude measures used in
the research.

In addition to highlighting the need for attitude assessment tools the reviewed
empirical literature provides a glimpse of the individual characteristics of students that
may inﬂuence the nature and strength of their attitudes toward the poor and their beliefs
about poverty. The following section details the descriptive variables included in the
personal characteristics and background portion of the survey used in the current study.

Socio-demographic correlates

The reviewed literature informed selection of relevant socio-demographic
variables for inclusion in the current study. The ﬁndings described above also shaped a
series of research questions to direct exploration of the data. No clear theories emerged
from the review of the literature on social work student attitudes. Therefore, testing of
theory-based hypotheses was not relevant to the current study. The socio-demographic
variables were examined to determine the nature and extent of relationships between
attitudes toward poverty and the poor and individual student characteristics. Potentially
relevant constructs reﬂected the psycho-socio-cultural environment of individual
respondents. The speciﬁc variables, which had been identiﬁed from the literature review,

included age, ethnicity and race, gender, socio-economic status (Atherton, et al., 1993;

35

Cryns, 1977; Grimm & Orten, 1973; Orten 1979, 1981), personal receipt of public
assistance, religion (Atherton, et al., 1993; Lowenberg, Dolgoff & Harrington, 2000;
Tropman, 1995), political orientation (Sowell, 1987), experience with the poor through
family or work (Beigi, 1998; Mofﬁc, Brochstein, Blattstein, & Adams, 2000), social
work experience (Hayes & Varley, 1965; Sharwell, 1974; Varley, 1963), intentions to
work with the poor (Reeser, 1984) and inﬂuences on the decision to enter the social work
profession.
Purpose of the Study
The current study was designed as a replication of the Attitudes Toward Poverty
scale development research. The purpose of the study, described below, was two-fold:
(1) to replicate the Atherton, et a1. scale development study using a sample of
MSW students and compare the replication sample data to the Atherton, et al.
sample.
(2) to expand on the Atherton, et al. research by examining the relationships of
attitude toward poverty and the poor with
a. traditional demographic characteristics (age, socioeconomic status and
background, family history of poverty, political afﬁliation, religious
heritage and afﬁliation),

b. various life experiences (personal receipt of government assistance,
informal, personal contacts with, and formal exposure to, the poor) and,

c. speciﬁc professional interests, inﬂuences and motivations.
The scale replication and comparison also included item content enhancements.
Additional items were included in the survey to assess more reﬁned dimensions of
poverty attitudes. New items also included statements about federal and state assistance

program policies that emerged after 1993.

36

Descriptive exploration was organized around student attitudes with respect to

student socio-demographic characteristics. Six conceptual variables provided structure

for the analysis of individual survey questions. These six concept areas were enumerated

as follows:

1.

6.

Dimensions of social inequality and oppression, (age, race, class, gender,
individual and familial history of poverty, and familiarity and contact with poor

peeple),
History of public assistance (personal use of governmental support),

Professional interests (plans to work with poverty or the poor within social work
profession, intended curriculum specialization),

Political orientation (political party afﬁliation, voting pattern, liberal/conservative
worldview, and inﬂuence of worldview on attitudes toward poverty and on
decision to become a social worker),

Religion and spirituality (spiritual/religious heritage, current religious afﬁliation,
inﬂuence of religion on attitudes toward poverty and on decision to become a

social worker),

Geographic effects (campus of study — main campus / distance education site).

Speciﬁc research questions were operationalized versions of these six conceptual
variables.

Research Questions

A number of questions about the relationship of attitude scores to demographic

and other personal variables were tested. Six exploratory questions specify that some

correlation or group differences will exist for speciﬁc variables but no statements are

made regarding the nature of those correlations or differences. In this study the non-

directional statements regarding attitude toward poverty and the poor were as follows:

Attitude scores may -

(1) be related to age.

(2) differ between students with and without race (“White”) privilege.

37

(3) differ between male and female students.
(4) differ among members of speciﬁc religions or spiritual traditions.
(5) differ among social work study specializations.

(6) be related to degree of change in social class or economic status from family
of origin to just prior to beginning MSW education.

Attitudes toward poverty (ATP) scale scores were such that higher scores
indicated more positive attitudes toward poverty and the poor. A number of questions
speciﬁed that some correlations or group differences may exist for particular variables
and the questions indicate an anticipated nature - or direction - of the relationships or
differences. Directional statements were generated for the following variables. More
positive attitudes may be found for respondents:

(7) with lower socioeconomic class scores for pre-MSW status and family of origin

(8) who indicate they were now or had ever been poor or that their parents or
grandparents had grownup poor or that they have known someone who was poor.

(9) with personal experience in receiving various types of government assistance and
for those with more extensive involvement with government assistance.

(10) expressing greater interest and reporting giving more thought to working with
the poor as professional social workers.

(11) who are afﬁliated with the Democratic party as opposed to Republican, or rated
their political orientation as more Liberal than Conservative.

(12) for whom their religion or spiritual tradition had greater inﬂuence on their
decision to become a social worker and on their attitudes about poverty and the
poon

(13) for whom their political orientation as a liberal or conservative had greater
inﬂuence on their decision to become a social worker.

(14) taking classes in more nrral, impoverished areas of the state.

38

In addition to statements about overall attitude scores, several others were proposed with
regard to individual scale items.

(15) Respondents who had been recipients of food stamps, unemployment
compensation, or Aid to Families with Dependent Children and General
Assistance, were expected to have more positive attitude ratings for speciﬁc
scale items relating to the corresponding assistance program.

Atherton, et al. (1993) examined attitude differences by gender and racial or
ethnic group. In the current study, the absolute number of students of color and male
students was so low that these items, in combination with each other or with other
demographic variables, could potentially identify students from these groups. In order to
protect student identities, data on these characteristics were not analyzed if the
combination led to a subgroup of the sample with less than ﬁve individuals.

Summary

This chapter offered a brief review of pertinent theoretical literature on values and
attitude measurement, provided a chronological review of research on social work student
attitudes about poverty and the poor, included a brief critique of gaps in the student
attitudes research, and posed a number of research questions to be explored after the
replication, or revision, of the ATP scale introduced by Atherton, et al. (1993). The next
chapter details the methodology used to replicate and explore the ATP scale results. The

research design, sampling procedure, sample characteristics, survey content, and data

analysis strategy are described.

39

CHAPTER 3
METHODOLOGY
Sample selection and characteristics

Graduate social work students were recruited through the 2004 fall semester
social work theory course on organization and community theories of human behavior in
the social environment (HBSE). Four class sections were approached to participate in the
survey of attitudes toward poverty and the poor. The course is required of all ﬁrst-year
Master’s level students on the main campus of a Midwestern university, and, during the
2004-2005 academic school year, for distance education students in other regions of the
state. This course provided access to nearly all newly enrolled MSW students and also
allowed for access to the distance education students in an outlying region of the state.
Outlying areas of the state are more rural and more economically depressed as compared
to Central and Southeast areas of the state. This geographic/economic heterogeneity is
valuable because it has the potential for resulting in greater sample variability.

Four class sections of the course were available for recruiting potential
participants for the study. The survey was completed by 61 ﬁrst year graduate students in
three course sections meeting on the main campus of the university and by another nine
(n=9) distance education students attending class in outlying areas of the state by means
of interactive television. These four class sections yielded a total study sample with 70
respondents.

Procedures
Instructors for each of the four HBSE course sections were asked to support the

study by sharing 30 minutes of class time for distribution of the survey and verbal

40

description of the project by the investigator. Students were given a study disclosure and
informed consent form attached to a 5-page survey and received verbally administered
instructions. The study disclosure and informed consent form included a summary of the
study, the anticipated risks and potential beneﬁts, and an estimate of the time needed to
complete the survey. The disclosure and consent form also contained a declaration of
conﬁdentiality and a statement about the voluntary nature of participation. Participants
received a duplicate copy of the informed consent forms with the names of contact people
available to them for support subsequent to participation. A copy of this form is included
in Appendix A.

Students were given 20 minutes to ﬁll out the ﬁve-page questionnaire (see
Appendix). The investigator distributed the survey and remained in the room to answer
questions and discuss the project further. At the end of the 20-minute period, the surveys
were collected. Surveys were collected without unique identiﬁers. Students were told
that once they handed in a completed survey, there would be no way for them to
withdraw from the research because their survey would not be identiﬁable (see
Appendix).

Signed informed consent forms were separated from the surveys as they were
collected. The informed consent forms were placed in an envelope, sealed and the
envelope ﬂap was signed as a means of creating a tamper-evident seal. This method of
collecting and storing informed consent forms was used to instill conﬁdence among
students that anonymity of the completed surveys would be maintained. The method also
met the university Institutional Review Board requirements for protecting human subjects

of research.

41

Measures

The survey instruments used for this study included 54 items of an attitude scale
and 22 additional questions regarding students’ demographic characteristics and
backgrounds. The corresponding variables are discussed in the following subsections.
The questionnaire was ﬁve pages long, single-sided. The attitude scale items were
presented on the ﬁrst two pages followed by the twenty-one descriptive questions (see
Appendix B).

Attitudes Toward Poverty and the Poor

The original 37 items of the Attitudes Toward Poverty (ATP) Scale were used as
written by the developers. (see Table l). Atherton, et al. (1993) reported strong internal
reliability for the scale (Cronbach’s alpha = .93; Split-half reliability = .87) and review of
the items suggested acceptable face and content validity. Items were rated on a ﬁve-
point Likert-type scale of agreement anchored with “Strongly Agree” and “Strongly
Disagree”. The rating scale was presented to respondents as had been used by the scale
developers. Atherton, et al. originally had offered respondents answer choices ordered as
follows: Strongly agree; Agree; Neutral on the item; Disagree; Strongly Disagree.
However, for the current study these choices were presented in a different order, that is:
Strongly Disagree; Disagree; Equally Agree and Disagree; Agree; Strongly Agree. Scale
anchor reordering of answer choices was used for ease of data coding and data entry.
Before data analysis began eight items were recoded to reverse the score per the scoring
instructions published by the developers, items #6, #11, #12, #22, #24, #25, #31, #36.

An additional 17 statements were included, immediately following the original 37

items. These items were created to capture more contemporary public welfare issues and

42

policies. The full text of all 54 attitudinal statements appears in Table 1 below and also

can be found in the Appendix. Two of the new items required reverse scoring, item #40

and item #43.

Table 1

Attitudes Toward Poverty and the Poor - original and new scale items

 

Original items

PWFQPPPPT‘

NNMNNNNNNN———-—-—-—————
soooxrchmpr—rcfoooggxmbww—p

A person receiving welfare should not have a nicer car than I do.
Poor people will remain poor regardless of what's done for them.
Welfare makes people lazy.

Any person can get ahead in this country.

Poor people are satisﬁed receiving beneﬁts.

Welfare recipients should be able to spend their money as they choose.

An able-bodied person using food stamps is ripping off the system.
Poor people are dishonest.

If poor people worked harder, they could escape poverty.

Most poor people are members of a minority group.

- People are poor due to circumstances beyond their control.

- Society has the responsibility to help poor people.

- People on welfare should be made to work for their beneﬁts.
- Unemployed poor people could ﬁnd jobs if they tried harder.
- Poor people are different from the rest of society.

Being poor is a choice.
Most poor people are satisﬁed with their standard of living.

- Poor people think they deserve to be supported.

Welfare mothers have babies to get more money.

- Children raised on welfare will never amount to anything.
- Poor people act differently.

- Poor people are discriminated against.

- Most poor people are dirty.

People who are poor should not be blamed for their misfortune.

- If I were poor, I would accept welfare beneﬁts.

- Out-of-work people ought to have to take the ﬁrst job that is offered.

- The government spends too much money on poverty programs.

- Some "poor" people live better than I do, considering all their beneﬁts.

- There is a lot of fraud among welfare recipients.

43

TCVCI’SC

30.

Beneﬁts for poor people consume a major part of the federal budget.

 

 

31- Poor people use food stamps wisely.
32- Poor people generally have lower intelligence than nonpoor people.
33- Poor people should be more closely supervised.
34- I believe poor people have a different set of values than do other people.
35- I believe poor people create their own difﬁculties.
36- I believe I could trust a poor person in my employ.
37. I would support a program that resulted in higher taxes to support social
programs for poor people.
New items ﬁadded
33- Welfare discourages people from working.
39. Poor people give up looking for work too quickly.
40- Poor people are no different from me except in life circumstances.
4]. Poor people are more responsible for their condition than many people would
like to admit.
42. It is unfair that welfare mothers get more money if they have a baby since
working parents who have another child don’t receive more money.
43- I believe people are poor because the US. economy excludes them.
44- Poor people don’t care about their neighborhoods.
45- Poor people don’t keep up the properties they rent or own.
46 A person going on welfare should have to submit to urine or blood tests for
' drugs.
47. I would feel uncomfortable visiting the home of a poor person.
48 I would support an increase in the sales tax to support social programs for poor
' people.
49 I would support an increase in the state income tax to support social program
' for poor people.
50. I would support an increase in the federal income tax to support social program
for poor people.
51- Many poor people do work but they get paid “under the table”.
52. The government should be able to limit the ways welfare and food stamp
recipients spend their beneﬁts.
53. Society should expect poor people to work in order to receive beneﬁts.
54. More and better substance abuse treatment would result in substantially less

*

poverty.

Indicates items that require reverse scoring

 

Note: Answer choices direct respondents as follows: "If you xxxxx, please circle XX" where xxxxx is,

Strongly Agree, Agree, are Neutral on the item, Disagree, Strongly Disagree, and XX is SA A N D SD

44

Three of the new items were a reﬁnement of an item from the original survey
regarding a respondent’s agreement with the statement “I would be willing to pay higher
taxes to support programs for the poor.” These three additional statements differentiated
between federal income taxes, state income taxes, and sales tax.

Personal Characteristics and Background

Unlike the 1993 Atherton, et a1. study, respondents were asked about their
personal characteristics and background. The directions given were, “for each of the
following questions, please choose the response that most closely describes you. If none
of the answer choices listed are acceptable or accurate, feel free to write in your preferred
response.” Questions were related to the following variables:

0 Age

0 Gender

0 Race and ethnic self-description as a measure of dichotomous race privilege

0 Social class or economic status and social class history (including class
mobility, familial intergenerational experience of poverty, personal
experience of poverty, personal familiarity with someone poor, and
employment or educational experiences with poor people).

0 Political identiﬁcation (including political party aﬂiliation, registration to
vote, frequency of voting, and self-description along a continuum of
liberal/conservative orientation).

0 Religious/spiritual heritage and current religious/spiritual afﬁliation

0 Personal history of receiving ﬁnancial or concrete assistance.

45

o Intended specialization within the social work curriculum, i.e. clinical or

macro practice.

0 Professional interests related to working with or for poverty and the poor.

0 Extent of inﬂuence of religion and political orientation on decision to become

a social worker and on attitudes toward poverty and the poor.

Five of the above variables were assessed using a modiﬁed magnitude estimation
measurement technique. Magnitude estimation is a measurement approach for generating
variables with an interval/ratio level of measurement (Lodge, 1981). Interval/ratio level
variables can be analyzed with more powerful parametric statistics (Pearson r, Student’s
t, ANOVA, linear regression). The lines, which represent a conceptual continuum, were
produced using an inkjet printer. The distance, in millimeters, between the left end of
each line and the mark made by each respondent was measured using a plastic ruler
purchased as an ofﬁce supply store. The measurement became a respondent’s score for
each variable. Lines were 180 millimeters long — an arbitrary length that resulted from
the margin settings on the survey pages.

Respondents were directed to mark a continuum, represented by a solid line, with
an X at the place that best described each of the following personal characteristics:

0 Respondent’s most recent social class/economic status and family of origin;

0 amount of thought given to working with or for the poor;

0 political orientation, varying between Far Left, Liberal, Conservative, and Far

Right; and

46

0 amount of inﬂuence religion/spiritual tradition and political orientation had on
their decisions to become social workers and on their attitudes toward the
poor

Data Analysis

This study was designed, in part, to be a replication of Atherton, et a1. (1993),
therefore, analysis initially followed that used by the scale developers. All analyses were
made using the Statistical Package for the Social Sciences (SPSS version 11.5).

Four of the 70 respondents were missing two scale item answers. A conservative
form of mean item substitution was used to replace the missing values. Sample means
were calculated for each of the four affected items and then used to replace the respective
missing answer. Atherton, et al. also had used mean substitution for a small number of
surveys. However, they chose to substitute a less conservative estimate of the mean
response to the individual respondent’s non-missing answers from the other 36 items.
The mean substitution method used for the current study was selected because it would
not alter the sample mean for items with missing values but it would allow for including
all participant surveys in all statistical analyses. The number of surveys with missing
scale item answers was so small as to be negligible in either study. Given this, it is
unlikely that either substitution method would affect scale metric properties and analyses.

Atherton, et al. had calculated discriminant power for each item as a secondary
method of reducing their original item pool from 50 to 37 items. Replication of item
reduction was not possible since the 13 items trimmed from the original scale were not
reported. Item reduction was not the purpose of the current research, therefore

discriminant power was not calculated.

47

Descriptive statistics were compiled by item, for the original scale, for a 54-item
scale, and for several subscales. Internal scale reliability was examined using both
Cronbach’s alpha and a split-half reliability procedure. Reliability analysis consisted of
computing item-to-total correlations.

A principle component factor analysis also was run using both orthogonal and
oblique rotation. Atherton, et al. (1993) reported that no factor structure emerged. For
the current study’s sample, factors were not anticipated for either the original 37 items or
a reﬁned version based on 54 items, but factor analysis was conducted to conﬁrm the lack
of factors.

Descriptive statistics for demographic questions were calculated. Variables
having an ordinal or interval level of measurement were correlated with the 37—item ATP
scale score, the larger, reﬁned scale, and each of the three conceptual subscales. Pearson
r coefﬁcients were computed to evaluate the statistical signiﬁcance and strength of
variable relationships of interest in the current study. Categorical questions, those having
a nominal level of measurement, having enough variability to justify analysis of group
differences were analyzed with respect to the ﬁve ATP Scale scores. Student’s t-test for
two independent samples and oneway analysis of variance (ANOVA) statistics for three
or more categories were used to examine differences among demographic subgroups.

For all statistical tests a probability level of .05 was considered statistically
signiﬁcant. For non-directional research questions, two-tailed tests of signiﬁcance were
calculated. Research questions, that speciﬁed the nature of a variable’s relationship to or

differences on the ATP scale score, were evaluated with one-tailed signiﬁcance tests.

48

CHAPTER 4
RESULTS

The primary purpose of this study was to replicate the work of Atherton, et al.
(1993). The secondary intent of the research was to provide an exploratory,
descriptive analysis of poverty attitudes with regard to MSW student demographics and
personal background characteristics. Exploration of the poverty attitudes scale was
guided by the research questions posed at the end of Chapter 2 — Review of the
Literature. ATP scale analyses are presented in the ﬁrst section, followed by
descriptive sections structured around the overarching concepts underlying the research
questions. Frequency analysis for variable distributions and appropriate descriptive
statistics are fumished to provide a sense of the homo/heterogeneity of the study sample
with respect to socio-demographic variables. Analyses of the poverty attitude scale by
socio-demographic and other individual characteristics follow each section of variable
descriptions.

Analysis of Attitudes Toward Poverty and the Poor Scale

Initially, analysis of the survey instrument mimicked the analytic procedures
reported by the original scale developers, Atherton et al. Additional analyses were
made to incorporate the 17 new statements added to the original 37 survey items. A
conceptual content analysis was used to create subscales, which were subsequently
reﬁned using empirical reliability analysis to craft internally reliable scales. Scale
distributions and inter-scale correlations are presented.

Individual ATP Scale Item Analysis
Examination of the means and standard deviations for each of the 54 scale items

allows for determination of the magnitude of item variability within any given sample.

49

Inspection of item means and variability allowed for identiﬁcation of survey questions
that do not exhibit sensitivity to attitudinal differences among respondents. Table 2
contains a list of item means and standard deviations sorted from widest to more

restricted variability. Restricted variability would indicate less sensitive items.

Table 2

Descriptive statistics for individual ATP items sorted by descending variability.

 

Item Item Std Skew Kur Mini Maxi
Num Content Dev Mean ness tosis mum mum

 

46 A person going on welfare should have to
submit to urine or blood tests for drugs.

6* Welfare recipients should be able to spend
their money as they choose. 1.1 2.8 0.32 -1.02 1 5

4 Any person can get ahead in this country. 1.1 3.6 -0.73 -0.26 1 5
10 Most poor people are members of a

1.3 3.5 -0.46 -0.98 l 5

minority group. 1.1 3.4 -0.49 -0.86 1. 5
26 Out-of-work people ought to have to take
the ﬁrst job that is offered. 1.1 3.7 -1.03 0.61 1 5

52 The government should be able to limit the
ways welfare and food stamp recipients

spend their beneﬁts. 1.1 3.1 0.31 -1.or 1 5
1 A person receiving welfare should not have

a nicer car than I do. 1.0 2.9 0.28 -0.40 1 S
13 People on welfare should be made to work

for their beneﬁts. 1.0 2.9 0.32 -0.55 1 5
21 Poor people act differently. 1.0 3.9 -0.65 -0.50 2
29 There is a lot of fraud among welfare

recipients. 1.0 3.5 -0.34 -0.87 2 5
32 Poor people generally have lower

intelligence than nonpoor people 1.0 4.0 -0.87 -0.28 2 5
47 I would feel uncomfortable visiting the

home of a poor person. 1.0 4.0 -1.28 1.14 1 5

48 I would support an increase in the sales tax

to support social programs for poor people. 1.0 2.1 0.95 0.40 1 5
51 Many poor people do work but they get

paid “under the table”. 1.0 3.2 0.05 -0.66 1 5

53 Society should expect poor people to work
in order to receive beneﬁts. 1.0 3.1 0.03 -0.66 1 5

50

54

22*
24*

5

15

18

28

30

34

37

4o

42

45

49

50

25*

11*

19

33

More and better substance abuse treatment

would result in substantially less poverty. 1.0
Poor people are discriminated against. 0.9
People who are poor should not be blamed

for their misfortune. 0.9
Poor people are satisﬁed receiving

beneﬁts. 0.9
Poor people are different from the rest of
society. 0.9
Poor people think they deserve to be
supported. 0.9
Some "poor" people live better than I do,
considering all their beneﬁts. 0.9
Beneﬁts for poor people consume a major

part of the federal budget. 0.9
I believe poor people have a different set of
values than do other people. 0.9

I would support a program that resulted in
higher taxes to support social programs for

poor people. 0.9
Poor people are no different from me
except in life circumstances. 0.9

It is unfair that welfare mothers get more
money if they have a baby since working
parents who have another child don’t

receive more money. 0.9
Poor people don’t keep up the properties
they rent or own. 0.9

I would support an increase in the state
income tax to support social program for

poor people. 0.9
I would support an increase in the federal
income tax to support social program for

poor people. 0.9
If I were poor, I would accept welfare

beneﬁts. 0.8
People are poor due to circumstances

beyond their control. 0.8
Welfare makes people lazy. 0.8
If poor people worked harder, they could
escape poverty. 0.8
Welfare mothers have babies to get more
money. 0.8
Poor pe0ple should be more closely
supervised. 0.8

51

2.7
4.3

3.7

4.0

3.5

3.9

4.1

4.0

2.0

4.0

3.6

3.7

2.0

2.0

3.9

3.6
4.1

4.1

4.1

4.0

0.27
-1.84

-0.46

-0.92

-0.67

-0.56

-1.10

-0.79

-0.68

0.79

-1.38

-0.61

-0.52

0.90

0.98

-0.62

-0.46
-1.05

-0.75

-l.05

-0.86

-0.55
3 .72

-0.41

0.97

0.08

-0.14

0.97

0.08

-0.15

0.24

2.84

-0.23

-0.52

1.20

1.22

0.43

-0.24
l .41

0.43

1.74

1.06

35 I believe poor people create their own
difﬁculties. 0.8 4.2 -0.69 0.36 2 5
38 Welfare discourages people from working. 0.8 3.8 -1.15 1.68 1 5

41 Poor people are more responsible for their
condition than many people would like to

admit. 0.8 3.8 -0.83 0.79 2 5
44 Poor people don’t care about their
neighborhoods. 0.8 4.1 -0.96 0.95 2 5
31 * Poor people use food stamps wisely. 0.7 3.3 -0.24 -0.49 2 5
2 Poor people will remain poor regardless of
what's done for them. 0.7 4.4 -1.16 2.31 2 5
l4 Unemployed poor people could ﬁnd jobs if
they tried harder. 0.7 4.0 -0.85 2.67 1 5
23 Most poor people are dirty. 0.7 4.2 -0.82 1.09 2 5
27 The government spends too much money
on poverty programs. 0.7 4.4 -1.08 1.01 2 5
39 Poor people give up looking for work too
quickly. 0.7 3.8 -0.57 0.50 2 5
43 I believe people are poor because the US.
economy excludes them. 0.7 3.9 -0.43 0.57 2 5
12* Society has the responsibility to help poor
people. 0.6 4.5 -1.00 -0.02 3 5
36* I believe I could trust a poor person in my
employ. 0.6 4.2 -0.49 1.12 2 5
7 An able-bodied person using food stamps
is ripping off the system. 0.6 4.2 -0.50 1.01
8 Poor people are dishonest. 0.6 4.6 -1.72 4.04 2 5
16 Being poor is a choice. 0.6 4.5 -1.19 2.07
17 Most poor people are satisﬁed with their
standard of living. 0.6 4.4 -0.52 -0.62 3 5
20 Children raised on welfare will never
amount to anything 0.6 4.6 -1.64 2.90 2 5

 

N0“: Sample n=70

Indicates items that were reverse coded before scale score creation
Values of skewness and kurtosis appearing in bold typeface indicate items
that are normally distributed.

Stereotypes underlie many of the survey items. For example, item #8 “Poor
people are dishonest.” Given prior research, which showed social work students have a
more positive attitude toward disadvantaged people than other professionals and the

public, the stereotypic nature of the items conceivably could polarize social work

52

respondent ratings or lead to restricted variability such that standard deviations and
minimum and maximum ratings would be smaller than exists for non-stereotypic items.
Review of items with the smallest variability supports this contention. Seven of the
lowest 20% of items (the last 11 statements listed in Table 2) are more stereotypic in
nature. Observed minimum and maximum ratings range between 2 and 5, and for one
statement, between 3 and 5.

Scale score creation and descriptive statistics for the original ATP scale

Scale scores for 70 respondents were calculated by summing each individual’s
responses, on a 1 to 5 point scale of agreement, across the original 37 survey items.
Potentially, scale scores can range from 37 to 185. Actual sample scores ranged from
110 to 174. The scale mean was 143.4 with a standard deviation of 13.2. The original
scale was normally distributed with acceptable values for skewness (.042) and kurtosis
(-.068). Just over two-thirds of students (68.3%) had scale scores between 130.2 and
156.6, the interval of values falling one standard deviation above and below the
observed mean.

The scale developers reported scale scores ranging from 79 to 177 for a sample
of 99 students attending social work and sociology courses at three different
universities. They reported a scale mean of 119.65 and standard deviation of 21 .97.
Scale scores were described as having a slight positive skew but otherwise normal
distribution. Exact values for skewness and kurtosis were not published. These
distribution characteristics indicate that the sample for the current study was more
homogeneous in their attitudes toward poverty and the poor than was noted in the study
reported by Atherton, et al. (1993). The discrepancy may be related to two differences

between the studies. First, the 1993 sample was drawn from both social work and

53

sociology courses, and from more than one university. These sampling differences
would have allowed for a greater range in opinions, both by academic disciplines and
by geographic areas. Second, the samples were drawn twelve years apart, the ﬁrst in
the early 1990’s and the second at the end of 2004. These sampling timeframes differ
substantially in dominant social policy and political climates of the two decades, as
well as differing by age cohorts.

ATP scale reliability and validity.
As noted earlier in this paper, review of the items suggested acceptable face and content
validity. Reliability analysis of the scale items yielded a Cronbach’s alpha of .87. This
coefﬁcient is at an acceptable level of internal reliability. Cronbach’s alpha for the
Atherton, et al. (1993) study was .93. The survey developers also calculated a
Guttman’s split-half reliability coefﬁcient of .87. The split-half coefﬁcient for this
study sample also was .87.

Item-to-total scale correlations are displayed in Table 3 along with the item-total
correlations reported by Atherton, et al. (1993). These correlations help to determine
whether any given item contributes uniquely to the underlying concept being measured,
in this case attitudes toward poverty and the poor. Relatively weaker correlations can
be indicative of an item’s contribution to the overall concept, providing it is not too
low. Very strong correlations should be interpreted to mean the item is not adding
much additional information to the underlying concept being assessed. Such items
presumably could be dropped from the scale without doing considerable damage to the
concept being validly measured. All items in the developers study exhibit acceptable
item-to-total correlations. Two items in the replication study, item 10 and item 15, had

low coefﬁcient values (.002, .177, respectively). However, item reduction was not one

54

of the intended functions of the study. These items were retained for creation of an
original scale score and used for remaining analyses.
Factor analysis.

Factor analysis is a statistical procedure used to detect an underlying factor
structure or subscales. Using the same principal components (PC) approach to factor
analysis employed by Atherton, eta1., factor analysis was carried out. Initial PC
analysis revealed 13 factors with eigenvalues greater than 1.0. However, all but 4 items
loaded on factor 1 with loadings of .30 or higher (the customary loading value in
deciding item assignment to factors. This parallels the factor analysis reported by
Atherton, et a1. (1993). They also found 10 factors with eigenvalues above 1.0. In their
study all 37 item loadings for factor 1 were .33 or greater. Visual inspection of the
scree plot for the replication sample led to the same conclusion reached by Atherton
and his colleagues. In both the original and current study factor eigenvalues declined
from a factor 1 value of 11.40 (Atherton, et al.) and 8.66 (current study) to factor 2
eigenvalues of 2.80 and 2.71 respectively. From these patterns it is apparent that no
empirical factor structure is inherent in the original poverty attitudes scale.

The factor analysis for the current study provided somewhat more detailed
information regarding the factor analysis of the scale items. The value of the Kaiser-
Meyer-Olin measure of sampling adequacy (KMO=.62) can be described as mediocre
(Kaiser, 1974). This statistical description of the 1993 study sampling adequacy was
not reported. In addition to the above characteristics noted in the factor analysis, the
orthogonal rotation of the factor matrix failed to converge in 25 iterations. Factor
rotation is used to reveal more simple structures across scale items. In this study

rotation was not possible. This ﬁnding is further assurance that empirical factoring is

55

not supported by the sample data. Because rotation was not possible in a customary 25
iterations the extracted factor matrix could not be reﬁned. The 37 items loadings on
factors remained difﬁcult to interpret because items had strong loadings on multiple
factors.

Factor analysis was abandoned in favor of a conceptual approach to identifying
subscales. A content analysis of all 54 items was completed. The results of the content
analysis are described in the next section, along with additional analyses of internal
reliability for conceptual subscales.

Content analysis of 5 4-item version of A T P scale for subscale creation

The 54 items that comprise the attitudes toward poverty and the poor scale in
the current study were reviewed and coded thematically by the investigator alone.
Investigator familiarity with literatures on attitudes research and public and
professional beliefs about social welfare informed the identiﬁcation of themes. No
other raters were used to validate the thematic coding. This procedural omission is a
study limitation and is discussed in the concluding chapter. Upon review, all 54 items
were statements regarding poverty to which respondents potentially could agree or
disagree. This review constituted conﬁrmation of the face validity of the scale.

The items also could be sorted into several categories. Several scale items were
related less to an individual’s opinion and more to respondent knowledge of facts about
poverty. For example, item 30 states that, “Beneﬁts for poor people consume a major
part of the federal budget.” If an individual is familiar with the Nation’s budget their
response would be based more on information than on personal opinion. In contrast,
item 27 taps an individual’s opinion about the expenditure of the nation’s funds, “The

government spends too much money on poverty programs.” This question can be

56

expected to yield an opinion-based response regardless of the respondent’s knowledge
about actual national budget ﬁgures. The ATP scale was content analyzed to separate
items about an individual’s own beliefs and feelings from statements that actually could
be answered with fact-based information about poverty or the poor. Initially the
categories were described as poverty beliefs, worldview, policy related, veriﬁable
statements, and personal preferences. Each of these categories were reﬁned using scale
reliability analysis, a customary approach to scaled variable creation. The initial and
ﬁnal items for each of the item content categories or themes, are listed in tables
following the narrative description of each category and relevant descriptive scale
statistics.
Poverty Beliefs

The largest number of items concerned beliefs that individuals hold about the
issue of poverty or about poor individuals, families, or communities. These items
included statements such as, “Welfare makes people lazy” and “Being poor is a
choice”. This category was initially comprised of 34 statements. The next largest
category contained items that were related to beliefs about the nature of the social
world or human nature. These 15 items were described as worldviews and included
such statements as, “Any person can get ahead in this country” and “Society has the
responsibility to help poor people”. The beliefs and worldview categories overlapped
extensively, they had 11 statements in common. These two initial item content
categories were ultimately blended, which resulted in a 38-item subscale. The 38 items
were tested for internal reliability and systematically reﬁned to create a single, 19-item
subscale (see Table 3). This subscale exhibited acceptable internal reliability, as

measured by Cronbach’s alpha. The scale alpha coefﬁcient was .83.

57

 

Table 3

Poverty beliefs subscale items and item-total correlations.

 

 

Item Item Content Item-total
Num Cronbach ’s Alpha = .83 correlations
3 Welfare makes people lazy. .37
4 Any person can get ahead in this country. .36
7 An able-bodied person using food stamps is ripping off the system. .29
8 Poor people are dishonest. .32
9 If poor pe0ple worked harder, they could escape poverty. .52
11 People are poor due to circumstances beyond their control. .45
12 Society has the responsibility to help poor people. .59
16 Being poor is a choice. .41
24 People who are poor should not be blamed for their misfortune. .49
31 Poor people use food stamps wisely. .36
34 I believe poor people have a different set of values than do other pe0ple. .43
35 I believe poor people create their own difﬁculties. .59
39 Poor people give up looking for work too quickly. .62
40 Poor people are no different from me except in life circumstances. .34
41 Poor people are more responsible for their condition than many people 53
would like to admit. '
It is unfair that welfare mothers get more money if they have a baby
42 since working parents who have another child don’t receive more .56
money.
43 I believe people are poor because the US. economy excludes them. .40
51 Many poor people do work but they get paid “under the table”. .18
54 More and better substance abuse treatment would result in substantially .21
less poverty.

 

Veriﬁable opinions
The next largest content area identiﬁed included 20 items that were statements
that have been or could be veriﬁed. Agreement with these statements constitutes a
personal opinion, but one which could be challenged or substantiated with factual
information. For example, “Beneﬁts for poor people consume a major part of the
federal budget” and “Poor people don’t care about their neighborhoods.” Such

statements can be conﬁrmed or disconﬁrrned with scientiﬁcally collected information.

58

Determining the veracity of the 20 statements is beyond the scope of the current study,
however the items shared a capacity for being tested through research. On this basis
they were combined to form a conceptual subscale. The items were tested for internal
reliability and combined to create a 20 item subscale (see Table 4). The subscale
exhibited acceptable internal reliability, as measured by Cronbach’s alpha. The scale
alpha coefﬁcient was .82.

Table 4

Veriﬁable opinions subscale items and item-total correlations.

 

 

 

Item Item Content Item-total
Num Cronbach ’3 Alpha = .82 correlations
2 Poor people will remain poor regardless of what's done for them. .35
5 Poor people are satisﬁed receiving beneﬁts. .54
10 Most poor people are members of a minority group. .05
14 Unemployed poor people could ﬁnd jobs if they tried harder. .39
15 Poor people are different from the rest of society. .23
17 Most poor people are satisﬁed with their standard of living. .57
18 Poor people think they deserve to be supported. .34
19 Welfare mothers have babies to get more money. .37
20 Children raised on welfare will never amount to anything. .46
21 Poor people act differently. .46
22 Poor people are discriminated against. .27
23 Most poor people are dirty. .52
27 The government spends too much money on poverty programs. .36
23 Some "poor" people live better than I do, considering all their beneﬁts. .39
29 There is a lot of fraud among welfare recipients. .52
30 Beneﬁts for poor people consume a major part of the federal budget. .58
32 Poor people generally have lower intelligence than nonpoor people .45
38 Welfare discourages people from working. .31
44 Poor people don’t care about their neighborhoods. .55
45 Poor peﬂle don’t keep up the properties they rent or own. .30

 

59

Policy Attitudes

A third type of item content was identiﬁed. Thirteen items were initially
selected as having policy related content. Respondents were asked their agreement
with items regarding social policy toward recipients of public assistance. Such items
included statements such as “Society should expect poor people to work” and “Poor
people should be more closely supervised”. These statements concern respondent
opinions about policies around poverty and recipients of public support. The original
13 policy related items were tested for internal reliability and systematically reﬁned to
create one, 9-item subscale (see Table 5). This subscale exhibited an acceptable level
of internal reliability, as measured by Cronbach’s alpha. The scale alpha coefﬁcient
was .81.

Table 5

Policy attitudes subscale items and item-total correlations.

 

 

Item Item Content 112?;
Num Cronbach ’s Alpha = .81 .
correlations
l A person receiving welfare should not have a nicer car than I do. .37
6 Welfare recipients should be able to spend their money as they choose. .45
13 People on welfare should be made to work for their beneﬁts. .69
26 Out-of-work people ought to have to take the ﬁrst job that is offered. .47
27 The government spends too much money on poverty programs. .19
33 Poor people should be more closely supervised. .42
46 A person going on welfare should have to submit to urine or blood 63
tests for drugs. '
52 The government should be able to limit the ways welfare and food .64

stamp recipients spend their beneﬁts.
53 Society should expect poor people to work in order to receive beneﬁts. .68

 

Personal Taxation Preferences
One additional distinction was noted through item content analysis. Seven

items were identiﬁed as containing attributions about oneself, i.e., the items contained a

60

reference to “I believe...” or “If I were poor. . .”. These “I” statements were combined
initially to create a seven item scale, which in turn was tested and reﬁned to create a
much smaller, 4 item subscale. This subscale exhibited an acceptable level of internal
reliability, with a Cronbach’s alpha coefﬁcient of .88. All four items were questions
related to personal preferences around paying higher taxes, of various forms, in order to
support social programs for the poor.

Table 6

Personal taxation preferences subscale items and item-total correlations.

 

 

Item Item Content Ittertna-l-
No. Cronbach ’s Alpha = .88 .0
correlations
“I would support “
7 a program that resulted in higher taxes to support social programs for poor 65
people. '

48 an increase in the sales tax to support social programs for poor people. .69
an increase in the state income tax to support social programs for poor 84
people. '

50 an increase in the federal income tax to support social programs for poor 82

people.

 

Inter-scale correlations
Another method for assessing the usefulness of creating subscales is inspecting
the extent of association between pairs of subscale scores. Pearson correlations were
calculated to assess the nature and strength of relationships, if any, among the
subscales. Correlation coefﬁcients are reported in Table 7. The correlation coefﬁcients
indicate, as would be expected, that the subscales share variance. This observation
makes sense because the original selection of items by Atherton, et al., was made to

measure a singular phenomenon, namely, attitude toward poverty and the poor.

61

However, no one correlation exceeds .68, indicating that the majority of variability

across the ATP items cannot be accounted for by just one of the subscales.

 

 

Table 7
Pearson correlation coefﬁcients and signiﬁcance levels for ATP subscales.
Poverty Policy Veriﬁable Tax
Subscales Belief Attitude Opinions Preferences
1 > .001 > .001 > .001
Poverty Belief
.64 1 > .001 > .001
Policy Attitude
Veriﬁable Opinions '68 '50 1 > '00]
Tax Preferences -.56 -.44 -.51 1

 

The basic descriptive statistics for each of the ﬁve scale scores, the original
ATP scale and the four subscales, are uninterpretable in relation to each other. This
difficulty is the result of scales containing a varying number of items. To address this
challenge, the scale scores were recomputed for purposes of comparison. Individual
scale scores were divided by the number of items that comprised each scale. For
example, for the original ATP scale, respondent scores were divided by 37 — the
number of items in the scale. This technique was used to recompute scores for each
scale, which subsequently can be understood with respect to the one to ﬁve point
Likert-type scale of agreement respondents used to rate each statement.

Descriptive statistics for each scale are presented in Table 8. All ﬁve scales are
normally distributed, based on Skewness and kurtosis values falling between —1 .00 and
+ 1.00. As indicated, score distributions can be interpreted relative to the 1 to 5 point
rating scale of agreement. Higher values, those closer to 5, reﬂect more positive
attitudes poverty and the poor with respect to the scale content. The most positive

attitude subscale is for veriﬁable statements about poverty or the poor, with a mean of

62

4.01 and a minimum observed rating of 3. 1. As described in the methodology, ratings
of 3 corresponded with an answer anchor of “Neutral on the item”. The next highest
scale mean is found with the original ATP scale, comprised of the 37 original Atherton,
et a1. items (M=3.88). The variability for the original scale (SD=.36) is comparatively
more restricted than the variability noted for the veriﬁable statements scale (SD=.41).
Minimum observed rating for this scale is 3.0.

Table 8

Descriptive statistics for the original ATP scale and four new subscales.

 

 

Scale variables Mean Std Dev Skewness Kurtosis Min Max
Original 37 item ATP scale 3'88 '36 -04 ' -07 3.0 4.7
Poverty Belief 3.86 .40 -.21 .34 2.8 4.8
Policy Attitude 3.38 .65 .16 -.47 2.0 5.0
Veriﬁable Opinion 4.01 .41 .08 -.26 3.1 5.0
2.03 .79 .40 -.26 1.0 4.0

Taxation preference

 

Note. Skewness & kurtosis values between — 1.0 and + 1.0 indicate a normal distribution

The poverty beliefs scale has a somewhat lower mean rating of 3.3 8, with a
standard deviation of 4.0, similar to that observed for the veriﬁable statements scale
(SD=.41). Minimum average score for the poverty beliefs scale (Min=2.8) is lower still
than the previous two scales. The policy attitudes scale has a scale mean of 3.3 8, a
value closer to the midpoint value of the response scale, i.e. 3, used for individual item
ratings. The standard deviation for the policy scale (SD=.65) is half again as large as
the variability seen on the poverty beliefs scale (SD=.40). The policy scale also has a
lower minimum observed score of 2.0. Finally, the lowest scale mean occurs for the

willingness to be taxed to support programs for the poor. This scale also has the

63

broadest variability with a standard deviation of .79 and a minimum observed scale
score of 1.0 (strongly agree) and a maximum of 4.0 (disagree). The taxation preference
scale was not recoded. Lower values indicate greater agreement with the four items
stated, “I would support an increase in ...”.

Content analysis of the 54 items in the attitudes toward poverty and the poor
scale resulted in one overall scale and four subscales for use in analysis of demographic
and background questions. The measures of internal reliability for the ﬁve scales
described above provided statistical reassurance that the thematic content identiﬁed
actually showed evidence of inter-item cohesion. The correlations among the subscales
indicate that the scales are related but are not so highly correlated as to be measuring
the same underlying variable. These qualities are desirable metric properties in the
conceptually created scales. The remainder of this chapter reports ﬁndings from the
analysis of respondent characteristics. Survey questions are grouped by the classes of
variables identiﬁed earlier in the dissertation.

Analysis of socio-demographics, background, and characteristics

Classes of variables provide the structure of the following sections. As noted in
Chapter 2, a series of research questions were posed. Questions were related to six
conceptual variable categories. The following table describes the six conceptual
variables explored in the remainder of this section. The table details the research
questions related to each conceptual category.

In the sections following Table 9, inferential tests of group differences and
relationships between groups are described in a narrative passage, followed by tables

summarizing the results of each statistical test.

64

Table 9

Research questions related to each of six major conceptual variables.

 

Dimensions of social inequaliﬂ

(1) be related to age.
(2) differ between students with and without race (“White”) privilege.
(3) differ between male and female students.

(6) be related to degree of change in social class or economic status from family of
origin to just prior to beginning MSW education.
(7) With lower socioeconomic class scores for pre-MSW status and family of origin

(8) who indicate they were now or had ever been poor or that their parents or
grandparents had grownup poor or that they have known someone who was poor.

History of public assistance

(9) with personal experience in receiving various types of government assistance and
for those with more extensive involvement with government assistance.

Professional interests

(10) differ among social work study specializations.

(1 l) expressing greater interest and reporting giving more thought to working with the
poor as professional social workers.

 

Political orientation

(1 1) who are affiliated with the Democratic party as opposed to Republican,
(1 1) rated their political orientation as more Liberal than Conservative.

(13) for whom their political orientation as a liberal or conservative had greater inﬂuence
on their decision to become a social worker.

(13) for whom their political orientation as a liberal or conservative had greater inﬂuence
on their poverty attitudes

 

Religion and spirituality

(4) differ among members of speciﬁc religions or spiritual traditions.

(12) for whom religion or spiritual tradition had greater inﬂuence on their decisions to
become social workers.

(12) for whom their religion or spiritual tradition had greater inﬂuence on their poverty
attitudes.

 

Gﬂgraphic effects

(14) taking classes in more rural, impoverished areas of the state.

 

65

Dimensions of social inequality

The absolute number of students of color and male students was very low.
Because these items, in combination with each other or with other demographic
variables, could potentially identify students from these groups, analysis was restricted
to just single dimensions. This limitation prevented conceptually more sound
comparisons that could have examined the intersecting effects of age, race, class, and
gender. However, for the sake of protecting conﬁdentiality, data on these
characteristics were not analyzed. The combination of demographic characteristics did
result in subgroups of students that would be identiﬁable as respondents.
Age

Respondent ages were not normally distributed (skewness and kurtosis values
between i 1.0 are considered normal.) The distribution of ages had a slight positive
skew (skewness=1.5) and was somewhat more peaked than a normal distribution
(kurtosis= 1.2). Ages ranged from 22 to 56. The mean age was 29.2 years with a
standard deviation was 9.1 years. Over half of the respondents (57.1%) were 25 years
old or less. Because no directional question had been made with regard to the
relationship between age and student attitudes, a 2-tailed pearson correlation coefﬁcient
was calculated. There were no signiﬁcant relationships between respondent age and
any of the poverty attitude scales scores.
Social advantage/disadvantage

The analysis of racial and ethnic categories is a troubling enterprise. Race is a
social construct, lacking a biological reality (Fields, 1987). However, race has been
used over time as a socially constructed dimension upon which social and economic

privileges are distributed throughout the United States. To the varying extents that

66

whiteness conveys race social advantages and non-white racial and ethnic groups are
disadvantaged, analysis of race/ethnicity information can be conceptualized as a
dichotomy of advantage and disadvantage. In this study, 75.7% of respondents (n=53)
self-described as White. The remaining 17 respondents constituted slightly less than
one-quarter (24.3%) of the sample. Because varying degrees of privilege are cenveyed
on the basis of skin color, respondents that self-reported a race/ethnicity category of
White, were analyzed as a group, in contrast to the group of respondents self-reporting
any non-White race/ethnicity. Any given individual’s experience of disadvantage and
privilege varies by unique life circumstances. However, non-White groups typically
are denied social or economic advantages on the basis of race or ethnicity. Two
students, counted among the disadvantaged subgroup, were international students who,
had they been born in the United States, would have been classiﬁed as disadvantaged.
Analysis of the social disadvantage variable was made using t-tests for
independent samples and a one-tailed test of mean group differences. Comparisons of
poverty scales revealed signiﬁcant differences between the groups on two of the four
subscales. There were no signiﬁcant differences between groups on attitudes toward
poverty policies or on preferences expressed regarding taxes. Signiﬁcant group
differences were noted on beliefs about poverty and veriﬁable statements or facts
regarding poverty. The mean poverty belief score for those with race privilege (White
respondents) was 74.3, and for those disadvantaged because of race/ethnicity the mean
was 70.5 (t=1.84, p < .05 one-tailed). The mean veriﬁable scale score for those with
race privilege was 81.3, versus a mean of 77.3 for those without such privilege (t=1.76,
p < .05, one-tailed). For both of the scales the White respondents appear to hold more

positive attitudes with respect to beliefs and potentially factual statements about

67

poverty and the poor. There is no observed difference in preferences for policies or
taxation related to poverty. These ﬁndings indicate that dichotomous race/ethnicity
based-groups do not differ in opinions about poverty solutions, but the groups do differ
in regard to beliefs about the causes and nature of poverty and with respect to
knowledge about veriﬁable poverty facts.
Gender

Similar to comparisons of race or ethnicity, gender presents a challenge to
interpretation of group comparisons. This challenge is particularly vexing when
comparisons involve social work students. The social work profession is
disproportionately inclusive of women (NASW, 2003). This discrepancy is quite large
with estimates typically around 80% female and 20% male. This difference in gender
proportions is present to an even greater extent in this study sample. Of the 70
respondents, 62 are female (88.6%) and 8 are male (11.4%). Such large discrepancies
in sample size between groups render detecting signiﬁcant mean differences extremely
unlikely. However, t-tests were run for each of the four scales. As expected group
differences were not statistically signiﬁcant for any of the poverty subscales.
Social class, class history & class mobility

The ﬁnal dimension of social inequality examined in the study involved
variables related to socio-economic status. Several measures of current socioeconomic
class and class history of each respondent’s family of origin were collected.
Respondents marked an anchored continuum to indicate their economic status or social
class just prior to starting the MSW program (Current) and also for their family of
origin (FOO). Continuum scores potentially could range from 0 (Poor/Lower Class) to

180 (Wealthy/Upper Class). Self-reported current socioeconomic class estimates

68

ranged from 18 to 167 with a mean of 80.4 and standard deviation of 30.2. Self-
reported family of origin socioeconomic class estimates ranged from 1 to 180 with a
mean of 79.2 and standard deviation of 39.6. The research question posed was about
whether students with lower SE class estimates would have more positive attitudes
toward poverty and the poor than those with higher social class estimates. Pearson
correlations between social class estimates and the poverty attitude score indicated
there was no signiﬁcant relationship between the four attitude scores and family of
origin-based social class estimates. For estimates of current social class a signiﬁcant
relationship was found for poverty beliefs, veriﬁable statements, and taxation
preferences. The correlation described weak negative relationships such that higher
social class estimates were related to less positive poverty belief scores (r = -.27), less
positive veriﬁable statement ratings (r = -.21), and less positive taxation preferences
(r = -.20). Each of the reported correlations was signiﬁcant at the .05 level (one-tailed).
There was no relationship between policy scale scores and current social class
estimates.

Class mobility was measured by calculating two change scores using self-
reported estimates of Current socioeconomic class and family of origin (FOO) social
class. The ﬁrst change score was a simple difference between the Current social class
estimate and the F00 estimate where FOO was subtracted from Current so that
downward mobility would appear numerically as a negative value. This measure of
change was directional; it allowed for testing the possibility that movement up the
economic mobility continuum would be related to less positive attitudes toward the

poor and likewise, movement down the continuum would be related to more personal

69

experiences with declining socio-economic status and more positive attitudes toward
the poor.

The second measure of social class change was the absolute value of the ﬁrst
change score. This non-directional difference score provided a measure of change that
disregarded the direction of change, either upward or downward, and indicated only
that a change had occurred. The reasoning behind this second measure of class
mobility was that substantial change from the social class or socio-economic status of
one’s family of origin, in either direction, may have a similar impact on attitude. That
is to say, ANY change may be more powerful in accounting for attitude than change in
one direction or another. This measure allows for the possibility that members of a
similar economic class level can hold negative attitudes about the poor. Also, the
second change score potentially corrects for some non-linear relationships. If the
scatter of attitude scores across the class change scores was non-linear and U shaped
the ﬁrst change score could appear to be uncorrelated with the attitude scale.

Directional change scores ranged from -89 to 145. Close inspection of the score
distribution revealed that one respondent was over three standard deviations
(SD=45.02) above the mean score of —1 .17. This respondent’s score was considered an
outlier and was dropped from the test of relationship strength between the class
mobility measure and attitude scores. Pearson r coefﬁcients calculated for the class
mobility measure paired with the poverty attitude scales indicated only one signiﬁcant
correlation between poverty beliefs scale and class mobility. The association between
the two was such that greater upward mobility was related to less positive attitude on
beliefs about poverty and the poor (r = - .24, p < .05, one-tailed). Relatedly, absolute

values for the change scores ranged from O to 145. After dropping the outlier value of

70

145, Pearson correlations between absolute change in social class and poverty attitude
scales were computed. No signiﬁcant correlations were found between attitude scores
and the simple, non-directional measure of change in social class/economic status.

Additional questions about respondents’ experiences with poverty and being
poor were asked. Respondents indicated whether they were now or had ever been poor,
whether they knew if their parents or grandparents had grownup poor and whether they
had ever personally known someone who was poor. Respondents chose among
responses of Yes, No or Don’t Know. Research questions posited that respondents
responding Yes to any of these questions would have more positive attitude scores.
Sixty-ﬁve students (n=65) indicated they knew or had known someone who was poor.
Three students had not personally known anyone poor, and two students indicated they
did not know whether they have known someone poor. Due to the lack of variability, no
tests were run on this variable.

For the remaining questions, about respondents’ own histories of being poor and
about poverty experienced by parents or grandparents, one-tailed Student’s t-tests for
independent samples were run to test for mean differences in attitude scores. Twenty
nine students (42.7%) indicated that they either currently were poor or had been poor in
the past. The survey questions did not specify a deﬁnition of “poor” or “poverty”
against which students could compare their own histories. Therefore, student responses
reﬂect entirely subjective, personalized deﬁnitions of what being “poor” meant to them.
Half the respondents (n=3 5) indicated they had not been poor (51.5%). Two
respondents left the question blank and four reported that they did not know whether
they had ever been poor. No signiﬁcant differences were revealed for any of the four

poverty scales. With respect to cross-generational poverty experiences, a large majority

71

of respondents (79.7%) described either or both their parents and grandparents as
having been poor (n=55). Another 15.9% indicated that neither parents nor
grandparents had been poor (n=11). Three respondents indicated they did not know
and one left the question blank. Comparison of the “yes” and “no” groups revealed no
signiﬁcant differences on any of the four poverty attitude scale variables.

Experience with poor people also can be garnered from employment,
internships, and volunteer work. When considering all three sources of contact with the
poor, 65 students (92.9%) described experience with the poor, and ﬁve (7.1%) had not
encountered the poor in any of the three forums. Here, again, the lack of variability
prevented comparisons on the basis of this variable. However, comparisons were
possible between having versus not having experience with the poor through
employment (44 vs. 21), through internships (43 vs. 22), and volunteer work (37 vs.
28). Despite more equivalent sample sizes, no signiﬁcant differences were found for
any of the three experience variables on any of the four poverty attitude scales.

Personal History Of Receiving Government Assistance

Respondents were asked to check mark from which of the listed government run
or subsidized programs they currently or had ever received assistance. Table 10 lists
the number and percentage of students indicating they participated each of ﬁfteen
public assistance programs.

As expected, a substantial majority (75.7 %) of respondents indicated they had
received ﬁnancial aid (n=53). Of the 15 programs listed, only ﬁve were endorsed by
enough respondents to justify further analysis; food stamps, unemployment

compensation, Medicaid, scholarships/fellowships, and federal ﬁnancial aid.

72

Table 10

Respondent history of participation in various social welfare programs

 

Forms of social welfare Signiﬁcant differences between

 

 

 

 

 

. Frequencies . . . .
assrstance recrprents and non-recrpients
Programs 3 percent Subscales
Beliefs Policies Veriﬁable Tax Prefs
AFDC 6 8.6%
General Assistance 5 7. 1%
Food Stamps 12 17.1% * ns * ns
881 3 4.3%
SSDI 2 2.9%
WIC 7 10.0%
Food Bank 4 5.7%
Scholarship or Fellowship 31 44.3% n5 "S nS "8
Homeless shelter 0 0.0%
Unemployment Compensation 12 17.1% ns * ns ns
Worker's compensation 3 4.3%
Medicaid 14 20.0% ns nS nS ns
Housing Assistance /Subsidy 2 2.9%
Veteran's Beneﬁts 2 2.9%
Federal Student Financial Aid 53 75.7% ns ns ns ns
Created assistance categories
Any government assistance
(excluding ﬁnancial aid & 29 41.4% ns ns ns ns
veteran's beneﬁts)
Received “welfare” 0 *
(ADC, GA, SSDI) 10 14.3 /0 ns ns ns
Entitlements
(workers compensation, 18 25.7% n5 ns ns ns

unemployment compensation,
SSI, & veteran’s beneﬁts)

 

Two additional categories, having a large enough 11 for analysis, were created by

combining recipients of ADC, General Assistance, and SSDI (n=10) into a Welfare

variable, and by combining program participants receiving Veterans’ Beneﬁts, SSI,

73

workers’ compensation, and unemployment compensation (n=18) into an Entitlement
variable. Finally, a category for students who reported having received any form of
assistance (other than ﬁnancial aid or veterans’ beneﬁts) was created. This category
contained 29 recipients of any of the various forms of government assistance.

Respondents with personal histories of receiving government assistance were
examined to address the research questions regarding experience with receiving public
assistance. The research question was regarding whether recipients of social welfare
programs would have more positive attitude scores. One-tailed t-tests for independent
samples were run to test for mean differences in attitude scores for the above categories
of assistance. A small number of signiﬁcant differences were found.
Food stamps

Twelve respondents (17.1%), who had received food stamps, had signiﬁcantly
less favorable beliefs (M=68.0) about poverty and the poor than the average beliefs
(M=74.5) reported by the remaining non-recipients (t=2.83, p < .05, two-tailed).
Likewise, the food stamp recipients (M=81.6) had signiﬁcantly lower subscale scores
about veriﬁable opinions than the average score (M=74.5) for the remaining non-
recipients (t=2.83, p < .05, two-tailed). No signiﬁcant differences were found between
food stamp recipients and non-recipients on agreement with policy related items or
willingness to pay more taxes.
Unemployment compensation

Twelve respondents (17.1%), who had drawn unemployment compensation at
some time, had signiﬁcantly greater agreement (M=32.9) for the policy subscale than

the average agreement (M=29.8) of non-recipients of unemployment beneﬁts (t=l .71, p

74

< .05, one-tailed). Unemployment recipients did not differ from non-recipients for any
of the three remaining subscales.
Welfare recipients

Ten respondents (14.3%), who had received either Aid to Dependent Children,
General Assistance, or Social Security Disability Insurance, had signiﬁcantly lower
agreement (M=68.7) on the poverty beliefs subscale than the average agreement of non-
recipients (M=74.2) on the poverty beliefs subscale (t=2.16, p < .05, two-tailed). No
other differences were found between welfare recipients and non-recipients on the other
poverty attitude subscales.
Other recipient categories

Several other recipient groups had high enough group sizes to allow comparison
with the corresponding non-recipient group. Comparisons on the four attitude
subscales were made between recipients and non-recipients of scholarships or
fellowships, Medicaid, federal student ﬁnancial aid, and entitlement programs
(speciﬁcally, veterans’ beneﬁts, Social Security, workers’ compensation, and
unemployment compensation). As indicated in Table 10 above, none of the
comparisons of groups’ differences were statistically signiﬁcant.
Receiving multiple forms of government assistance

An additional variable was created by making a simple count of the total
number of government programs from which a respondent reported having received
assistance, not including ﬁnancial aid. Education related aid was excluded from the
count because so many of the respondents (81.4%, n=57) reported having received
federal ﬁnancial aid, scholarships, or fellowships. The count variable provided a rough

measure of the extent of experience respondents had with government assistance. The

75

count ranged from 0 (n=40) to 7 (n=1), with a mean of 1.01 and a standard deviation of
1.60.

The relationship of exten t of government assistance with attitude subscales was
assessed using Pearson’s correlation. There were no signiﬁcant correlations between
the subscale variables and receipt of government assistance. A total of 29 respondents
indicated they participated in anywhere from 1 to 7 of the government assistance
programs listed, not including ﬁnancial aid. Thirteen of these 29 had only received
assistance through one program. A dichotomous variable was created to compare the
forty individuals who had never received government assistance (58.6%) with the
twenty-nine respondents who had received assistance from one or more government
programs (41.4%). Perhaps an experiential or attitudinal difference exists for people
once they become a “recipient”. The mean attitude score for this group of 29 was
compared with the remaining respondents using a t-test for independent samples. No
signiﬁcant differences between the two groups was found for any of the attitude
subscales.

In addition to the above comparisons, several other questions related to
individual scale items, were examined. Respondents, who had been recipients of food
stamps, unemployment compensation, or Aid to Dependent Children or general
assistance, were examined to reveal whether they had higher agreement ratings for the
speciﬁc scale items related to the corresponding assistance program. Table 11 lists the
speciﬁc items analyzed for recipient/non-recipient rating differences. The agreement
rating scale has an ordinal level of measurement; therefore, the non-parametric Mann-

Whitney U statistic for independent samples was used.

76

 

Table 11

Program recipient differences for scale items related to specific programs

 

 

 

 

 

Recipients of ATP item number examined sig
7. An able-bodied person using food stamps is ripping ns
Food Stainps off system.
31. Poor people use food stamps wisely. ns
14. Unemployed poor people could ﬁnd jobs if they tried ns
Unemployment harder
Compensation 26. Out of work people ought to take the ﬁrst job that is < .05*
offered
Welfare (ADC, 25. If I were poor I would accept welfare beneﬁts ns
GA, SSDI)

 

* Indicates greater disagreement by recipient group

Only one signiﬁcant difference was revealed. Individuals who had received
unemployment compensation disagreed (M rank = 45.1) signiﬁcantly more than non—
recipients (M rank = 32.9) with the policy related statement, “Out of work people ought
to have to take the ﬁrst job that is offered” (U = 220.5, p < .05).
Interests within professional social work

Respondents were asked three questions to determine their interests within the
profession. Professional interest in working with or for the poor was assessed with a 5
choice question. The numbers and percentages of students choosing each response
choice are displayed in Table 12. Mean differences in attitude subscale scores were
examined for respondents with no speciﬁc interest in working with/for the poor (n=17),
those who would not mind and those speciﬁcally seeking an MSW in order to work
with/for the poor. A oneway AN OVA was run. Signiﬁcant differences between
interest groups were found only for attitudes toward the veriﬁable, knowledge-based

items subscale. Differences among the three interest groups were such that the 17

77

students who had not given any thought to working with the poor, or had speciﬁc
interest in other working with other populations had signiﬁcantly lower (M=75.2)
veriﬁable subscale scores (F = 4.68, df (2,66), p < .05) than both the 35 students who
would not mind working with or for the poor (M=81.9) and the 17 students who
speciﬁcally sought an MSW in order to work with the poor (M=81.2). Tukey b, post
hoc analysis conﬁrmed speciﬁc mean differences between the groups.

Table 12

Type and extent of student interest in working with or for the poor

 

What professional interest do you have in working with poor people n percent
and/or with policies or programs for poor people?

 

 

Had not given it any thought 4 5.8%

No interest, I have a speciﬁc interest in other populations 13 18.8%

No interest, I would not / could not work with/for the poor 0 0.0%

Some interest, I would not mind working with/for the poor 35 50.7%

High interest, I speciﬁcally sought an MSW in order to work

with/for the poor 17 24.6%
Sample n=70, one respondent did not answer 69 100%

 

Respondents also indicated on an anchored continuum the amount of thought
they had given to working with poor people or policies and programs for the poor.
Continuum ratings could vary between 0 (None) and 180 (A great deal). Actual
responses ranged from O to 172 with a mean of 99.23 and standard deviation of 47.4.
Ratings had a nearly normal distribution with only slight negative kurtosis (~1.3); the
distribution was a bit ﬂatter than a typical bell-shaped distribution. One research
question posited that those respondents giving more thought to working with the poor
would have more positive attitude subscale scores. One-tailed Pearson correlation

coefﬁcients were calculated to assess these relationships.

78

Three signiﬁcant (p < .05) positive associations were revealed between amount
of thought given to working with/for the poor and the policy subscale (r = .29),
veriﬁable items subscale (r = .28), and taxation attitudes subscale scores (r = - .28). No
correlation existed between amount of thought given to working with the poor and the
general beliefs subscale. The more thought given to working with or for the poor, the
more positive the attitude scores for the policy subscale and for the veriﬁable subscale,
and the less positive the score on taxation subscale.

Respondents also were asked to identify which programmatic area of study they
were pursuing within the MSW curriculum. MSW students choose either a clinical
specialization or a macro practice specialization in Organizations, Communities, and
Policy (OCP). Two thirds of the respondents (n=46) identiﬁed the clinical
specialization. Twenty-one percent (n=15) identiﬁed OCP and thirteen percent (n=9)
were undecided. The research question posed with regard to differences for these
specializations was addressed using two-tailed t-tests for independent samples. The
tests revealed signiﬁcant group differences (t = 2.03, p < .05, two-tailed) only for
attitude scores on the policy subscale, such that, OCP (macro practice) students had
more positive policy subscale scores (M = 33.3) than reported by Clinical (micro
practice) students (M=29.73). No signiﬁcant differences were found for the beliefs,
veriﬁable, or taxation willingness subscales.

Political orientation or worldview

Several questions were used to identify respondent political orientations or view
of the social world, namely afﬁliations with various political parties, voting registration
status, voting frequency, and orientation to the social world along a continuum of

Liberal / Conservative. Continuum scores could range from 0 (Far Leﬁ) to 180 (Far

79

Right). Liberal and conservative rating anchors were spaced evenly along the
continuum leaving a middle segment of the rating line open for respondents to indicate
a middle of the road position.

A wide spectrum of party afﬁliations was endorsed by respondents. The two
largest groups were None (n=11, 15.7%) and Democrat (n=44; 62.9%). Smaller
numbers of students reported afﬁliations with other parties, speciﬁcally six respondents
were Independents (8.6%), ﬁve were Republicans (7.1%), and four indicated other
political parties (5.7%). A comparison of Democrats and Republicans had been
planned but was not done because too few students identiﬁed as Republican (n=5).

All but 4 students were registered to vote (n=66, 96.4%). However, there was a
broad range in the frequency with which they voted. One student indicated they never
voted (1.5%), 27 students vote only in the November presidential elections every four
years (41.5%), 16 only vote in local elections (24.6%), 5 students vote in yearly
primaries (7.7%), and 16 students indicated they vote regularly throughout each year
(24.6%). Correlations between voting frequency and attitudes subscales were
calculated using Spearman’s rho. No correlations were statistically signiﬁcant and all
coefﬁcients were very low.

The continuum scores for liberal / conservative worldview or political
orientation ranged from 0 to 145 (of a possible upper value of 180) with a mean of 72.5
and standard deviation of 31.2. The distribution was normally distributed. One
research question posited that respondents with a more Liberal orientation would have
more positive attitude scores. One-tailed Pearson correlation coefﬁcients were
calculated to assess relationships between worldview and the attitude subscales.

Signiﬁcant negative correlations (p < .05) were revealed between worldview and the

80

beliefs subscale (r = -.26), the policy subscale (r = -.27), the veriﬁable subscale (r = -
.33) and the taxation willingness subscale (r = .39) toward poverty and the poor such
that a more liberal orientation (lower score) was related to more positive attitudes
(higher subscale scores)

Respondents also indicated the extents to which their liberal or conservative
views inﬂuenced their decisions to become social workers and their attitudes toward
poverty and the poor. Inﬂuence of worldview on attitudes were estimated along a
continuum of inﬂuence ranging from 0, “no inﬂuence at all on my attitudes about
poverty and the poor”, to 180, “the greatest inﬂuence of all on my attitudes about
poverty and the poor”. Inﬂuence estimates along the continuum ranged from O to 175
with a mean of 107.5 and standard deviation of 37.4. The estimates were normally
distributed. One-tailed Pearson correlation coefﬁcients were calculated to assess
relationships between worldview inﬂuences on the attitude subscales. Signiﬁcant
positive correlations (p < .05) were revealed between inﬂuence of worldview on
poverty attitudes and the beliefs subscale (r = .39), the policy subscale (r = .40), the
veriﬁable subscale (r = .34) and the taxation willingness subscale (r = -.47) toward
poverty and the poor, such that the greater the inﬂuence of worldview on attitudes
about poverty (greater inﬂuence) the more positive the attitudes (higher subscale
scores).

It is interesting to note that the relationship between worldview and self-
reported inﬂuence of worldview on attitude is statistically signiﬁcant and relatively
pronounced (r = —.37, p < .05); the more liberal a respondent’s rating the greater the
reported inﬂuence of worldview on poverty attitudes. Likewise the relationship

between worldview and self-reported inﬂuence of worldview on decision to become a

81

social worker was statistically signiﬁcant and quite strong (r = -.51, p < .05). More
liberal ratings were associated with ratings for greater inﬂuence of worldview on
decision to become a social worker.

Relatedly, the relationship between worldview and self reported inﬂuences of
religion on attitudes toward the poor was statistically signiﬁcant (p < .05).
Respondents with more conservative ratings on worldview reported religion having a
greater inﬂuence on their attitudes about poverty and the poor (r = .36). More
conservative ratings on worldview also were associated with increased inﬂuence of
religion on decision to become a social worker (r = .28). These ﬁndings raise an
additional question about worldview. Do different types of current religious afﬁliations
and different religious heritages from family of origin differ in respondent worldview?
This question was addressed by comparing the three categories of religions (None,
Catholic, Protestant/Christian) on average worldview estimates. None of the
comparisons, for either current religion or family religion during childhood, were
statistically signiﬁcant.

Religious/Spiritual A ﬂiliation

Respondents were asked to specify with which of six religious/spiritual
traditions they were currently afﬁliated and indicate with which religious heritage they
had grown up. The sample currently was disproportionately Christian (n=51) —
including Catholic, Protestant, or Other Christian - (72.9%;). A small number (n=1 1)
indicated they had no religious afﬁliation currently (15.7%). The remaining eight
students described various other religious affiliations (11.4%).

With respect to religious afﬁliation of family of origin, the sample largely was

Christian (n=57) with 81.4% indicating they had grown up Catholic, Protestant, or

82

Other Christian. Only 9 students indicated they had grown up without a religion
(12.9%). A small proportion of other respondents (5.7%) did not answer the question
or speciﬁed other spiritual/religious heritages, e. g. Jewish, Native American spiritual
heritage, Hindi.

Respondents also were asked to indicate the extent to which their
religious/spiritual afﬁliation or heritage inﬂuenced their attitudes about poverty and the
poor and had inﬂuenced their decision to become a social worker. Respondents
answered by marking a continuum of inﬂuence which ranged from O (No inﬂuence at
all on “my attitudes about poverty”, “my decision to become a social worker”) to 180
(Greatest inﬂuence of all on “my attitudes about poverty”, “my decision to become a
social worker”). Both distribution of inﬂuence estimates ranged from 0 to 180 for
inﬂuence variables. Inﬂuence of religion on attitudes toward the poor had a mean of
81.3 and standard deviation of 50.0. Inﬂuence of religion on decision to become a
social worker had a mean of 78.4 and standard deviation of 54.8. Both distributions
were slightly platykurtic (-1.1), neither was skewed. The research questions were
examined with respect to inﬂuence of religion on attitudes about poverty and on
decision to be a social worker. In both instances the questions posited that for those
individuals reporting religion or spirituality had greater inﬂuence on their decision to
become a social worker, or on their attitudes about poverty, would have more positive
attitudes toward poverty and the poor. One-tailed Pearson correlation coefﬁcients
were calculated. There were no signiﬁcant relationships between each variable and the

four attitude subscales. The observed strengths of associations were uniformly weak

(no r>i .17).

83

Geographic effects

Finally, comparison had been planned for students living and attending school
in upstate regions through the school’s MSW distance education program (n=9) and
those attending the MSW program on the main campus (n=61). Respondents studying
in comparatively more rural areas may have had more positive attitude scores given that
exposure to poor people and poverty would be a more common experience and might
help to challenge stereotypes. A t-test for independent samples revealed only one
signiﬁcant difference on attitude subscales. This ﬁnding was an unexpected one given
the disproportionate numbers of respondents in the two groups (i.e. 9 versus 61).
Despite the sample size discrepancy, students in the distance education group had
signiﬁcantly lower, i.e. less positive, belief subscale scores (M = 68.7) than the group
of students enrolled at the main university campus (M=74.0) (t=2.0l , p < .05, two-
tailed). However, the group differences were in an opposite direction to that suggested
by the research question. The relevance of this discrepancy will require additional
research. No signiﬁcant differences were found for the other three subscales.

Findings Overview

The analysis results described above addressed the primary and secondary
purposes of the research. Evidence was reviewed regarding replication of the original
ATP scale and enhancements to the scale were detailed. The descriptive function of the
study was addressed through systematic exploration of demographic variable categories
with respect to the attitude subscales.

Chapter 5 provides a discussion of the exploratory ﬁndings. Discussion is

followed by conclusions, and a review of the implications of the study for the larger

84

issue of evaluating social work education. The limitations of the study are described in

detail. Directions for further research are suggested.

 

85

CHAPTER 5
SUMMARY, CONCLUSIONS AND IMPLICATIONS

This ﬁnal section of the report is divided into four topics. Discussion of the
analysis results will be organized around the two purposes of the study. Each core
purpose is discussed separately. Implications of the study results for social work
education follow the summary and discussion of ﬁndings. The chapter concludes with a
critique of the study’s limitations and suggestions for ﬁrture research.

The ﬁndings reported in the preceding chapter can be understood in a variety of
ways. This multitude of meanings illuminates the multi-faceted nature of psychosocial
phenomena like professionals’ attitudes toward poverty and the poor. The attitudinal data
reﬂect the complexities inherent in measuring attitudes and hypothesizing their effects on
professional comportment, i.e., demeanor and behavior.

Summary and discussion

The primary purpose of the study was to replicate the Atherton, et al. scale
development research (1993), including comparison of descriptive statistics, scale
reliability analysis, factor analysis, and inter-item correlations. The second study purpose
was to examine research questions about the relationships between attitude scores and:
traditional socio-demographic characteristics related to dimensions of social inequality;
religious afﬁliation, heritage and inﬂuences; political orientation, voting behavior,
worldview and inﬂuences; history of using government assistance; and professional
interests and social work specialization; and regional differences. Findings informing
each purpose are reviewed brieﬂy and discussed in the context of the evaluation logic

model presented in Chapter 1.

86

 

Replication

The 1993 Atherton, et al. ATP scale development was replicated successfully. A
brief critique of the scale development report is offered. As stated in Chapter 1 the
journal article that introduced the scale did not report individual item variances. It would
have been useful to inspect item variability and compare the ﬁgures for the two different
student samples — social work/sociology and business students — for the original study.
Access to the 1993 descriptive statistics also would have allowed for comparisons with
the MSW student sample used in this study.

This replication study had a lower, but still acceptable level of internal scale
reliability than was reported in 1993 by Atherton, et al. Item-to-total correlations also
were lower than those found by the scale developers. Item-to-total correlations in the
current research suggested that it may be possible to adequately measure attitudes about
poverty with fewer than the original 37 items, or 54 items as in this study. Shorter
surveys are less time consuming for respondents to answer, are less resource intensive to
code and analyze, and therefore, are preferable to longer surveys. Scale content can be
reduced further and with greater conﬁdence if restricted response ranges consistently are
observed across studies for particular items. Knowing individual item distribution
characteristics from multiple studies would have made possible comparison of item
means and variability. Until the ATP scale is used with more samples, no change or
reduction in items is warranted.

Factor analysis of the replication survey data produced results similar to those
reported by the scale developers. No factor structure is inherent in the ATP scale.
However, the factor rotation method used in both studies - Varimax - assumes that if

subscales exist they are orthogonal, that is to say, uncorrelated with each other. Given

87

 

the realities of poverty and the complexity of attitudes and beliefs, the assumption of
orthogonal factors may not be realistic. Future factor analysis should explore the
possibility of inter-correlated factors. An oblique rotation approach should be used to
detect factors.

The study results indicated factors could be constructed conceptually instead of
empirically. A literature-based argument can be made for a conceptual. approach to
subscale or factor identiﬁcation. The studies on student attitudes from the 1960’s and
1970’s, reported on by Sharwell (1974) and Orten (1976) and summarized in the review
of empirical studies, used a conceptual approach to measuring attitudes toward poverty
and the poor. From an evaluation standpoint, it is important to separate into subscales the
items of the ATP that potentially can be changed through social work education from
those statements that reﬂect a particular worldview about human nature or political
viewpoint about the proper role of government. Worldviews or political viewpoints may
be amenable to change but would require more and different effort than may be achieved
by presenting factual information acquired through research.

As mentioned in the study methodology, Chapter 3, item review suggested
acceptable face and content validity. However, items appeared to be based on gross
stereotypes or global attributions about “the poor” as a unitary group. The reviewed
literature clearly notes that social work students do have a more positive attitude toward
the disadvantaged, than either students from other disciplines or the general public. More
reﬁned distinctions in attitudes may go undetected by blanket statements or
oversimpliﬁed items.

There is reason to believe that self-reported attitudes toward speciﬁc

subpopulations of the poor, (e. g., single mothers, or unique individuals), speciﬁc cases

88

encountered in practice settings or through vignettes, may differ from espoused attitudes
which take as their referent the unitary construct of “the poor” (Tamoush, 1998). The
implications of group versus individual representations of the poor for measurement of
attitudes should be the focus of ongoing research.

At least 33 of the 54 scale items were worded to include all poor people or refer
to poor people as a homogeneous group. Tamoush (1998) indicates that the nature of the
referent of poverty attitudes, i.e., group versus individual stimuli, affects attitudes toward
poverty and the poor. The speciﬁcation of a poor individual in a given item, as opposed
to referring to the “poor” as a group, was related to different attitudes among the general
public Tamoush sampled. Tamoush addresses the historical and ongoing relevance of the
concepts of deserving and undeserving poor to public discussion. Future research should
consider the importance of deservingness of particular individuals or subgroups of the
poor for assessing attitudes among professional social workers and the profession, as a
whole.

Overall, the current study ﬁndings support use of the ATP scale as a valid and
reliable attitude assessment tool. To more comprehensively validate the instrument, the
ATP survey should be used in longitudinal studies with samples of social work students
and practicing social work professionals. The scale also should be tested with other
student groups and non-student populations such as poor people, working class
individuals, voters, and civil servants and other employees of public assistance programs.
In addition, the relationship of social work student and professional attitudes to actual
service delivery behavior and professional demeanor with disadvantaged clients should
be assessed. Attitudes can be somewhat predictive, but are not a perfect predictor or

cause, of behavior, professional or otherwise (Azj en & Fishbein, 1975). Knowing the

89

extent to which the ATP can predict professional behavior would be relevant for
establishing the scale’s scope of usefulness for evaluating social work educational

outcomes related to social work values.

Soda-demographics, personal characteristics and background variables

The second purpose of this study was exploratory and descriptive. Research
questions were posed to direct the examination of relationships between ATP sub scale
scores and a variety of socio-demographic, life experience, and professional interest
variables.

Table 13

Findings for non-directional research questions.

 

 

Attitude scores will:
(1) be related to age no signiﬁcant
correlation
(2) differ among social work study specializations. no signiﬁcant
differences
(3) differ among members of speciﬁc religions or spiritual no signiﬁcant
traditions differences
(4) be correlated with frequency of attendance at religious no signiﬁcant
services correlation
(5) be related to degree of change in social class or economic signiﬁcant,
status from family of origin to just prior to beginning MSW weak direct
education correlation

 

Nine research questions implied the nature and/or extent of relationships among
personal characteristics or background variables to attitudes. In Table 14, each question

is repeated and the corresponding result is indicated.

90

 

Table 14

Findings for directional research questions.

 

More positive attitudes will be reported by respondents:

 

(6) With lower socioeconomic class scores Signiﬁcant, weak correlation
from pre-MSW student status and family of between current social class and
origin poverty beliefs, no signiﬁcant

correlation with family of origin
social class

 

(7) Who indicated they were now or had ever No signiﬁcant differences
been poor or that their parents or
grandparents had grownup poor or that they
have known someone who was poor.

 

(8) with personal experience in receiving No signiﬁcant differences or
government assistance and for those with correlations
more extensive involvement with
government assistance.

 

 

(9) expressing speciﬁc interest and reporting No signiﬁcant differences for
giving more thought to working with the expressed interests in work with the
poor as professional social workers. poor. Signiﬁcant direct correlations

with policy attitudes, veriﬁable
opinions, tax preferences.

 

(10) who afﬁliated with the Democratic party as Political party affiliation had

opposed to Republican, or rated their restricted variability and was not
political orientation as more Liberal than analyzed. Signiﬁcant correlation
Conservative. between more Liberal worldview

and all subscales.

 

(l l) for religions, now or heritage from family of No signiﬁcant differences for

origin, for whom religion or spiritual current or family of origin religious
tradition had greater inﬂuence on their afﬁliations.

decision to become a social worker and on No signiﬁcant correlations between
their attitudes about poverty and the poor. any inﬂuences on social work

' decision and poverty attitudes.

 

(12) for whom their political orientation as a Signiﬁcant correlations on each
liberal or conservative had greater inﬂuence subscale with worldview inﬂuence
on their decision to become a social worker. on poverty attitudes and worldview

inﬂuence on social work decision.

 

(13) taking classes in more rural, impoverished Signiﬁcant difference between
areas of the state through a distance campus of study for poverty beliefs.
education program. No signiﬁcant differences on

remaining subscales.

 

91

In addition to the statements about attitude scale scores, several other questions
had been proposed. Respondents, who had received food stamps, unemployment
compensation, Aid to Dependent Children, general assistance or SSDI, were expected to
have more positive attitude ratings for speciﬁc scale items related to the corresponding
assistance program. Recipients of federal student ﬁnancial aid were expected to have
more positive ATP scale scores.

Table 15
Findings for differences between recipients and non-recipients on program-related A T P

scale items.

 

 

Number of
Recipients of Related ATP item Recipients Analysis
(non-recipients) results

 

Food Stamps 7. An able-bodied person using food
stamps is ripping off the system.

 

 

12 No signiﬁcant
d'ff

31. Poor people use food stamps wisely. (5 7) l ﬂames
Unemployment 14. Unemployed poor people could ﬁnd N0 Signiﬁcant
jobs if they tried harder differences

12
26. Out of work people ought to have to f 5 7) dii'ggrgfgzgt

take the ﬁrst job that is offered

AFDC, GA, & 25. If I were poor I would accept welfare 10 No signiﬁcant
SSDI beneﬁts (59) dlfferences
Financial Aid ATP scale score 50 No signiﬁcant
(20) differences

 

The majority of the original 15 questions were answered. Of the ones that
showed statistically signiﬁcant results, two ﬁndings were not surprising. The more

thought a student gives to working with the poor, the more positive their attitudes toward

92

the poor. The more Liberal a student identiﬁes as a worldview or in their political
orientation, the more positive their attitudes toward the poor.

Findings for two of the hypotheses were noteworthy. The non-directional
hypothesis that degree of class mobility would be related to attitude was conﬁrmed.
Students, who experienced greater downward mobility, indicated less positive attitudes
toward the poor. The experience of downward mobility may prejudice individuals about
poverty and the poor.

The second unexpected ﬁnding was that the 10 individuals who reported having
received Aid to Dependent Children, General Assistance, or Social Security Disability
Insurance d_ic_l_n_o_t express signiﬁcantly more agreement with the item “If I were poor I
would accept welfare beneﬁts.” Since these respondents had a history of accepting
welfare beneﬁts they reasonably could be expected to have more agreement with the
hypothetical situation posed by the statement. It is not uncommon for individuals, who
are members of, or fear becoming members of, a stigmatized group to hold negative or
stereotypic beliefs about other group members (Goffrnan, 1963; Miller, 1976; Crosby,
Pufall, Snyder, O’Connell & Whalen, 1989; Beigi, 1998). Explicit examination of
alternative explanations for the ﬁndings is necessary. And, replication of these ﬁndings,
and deeper investigation of their meanings, is needed before credible interpretation can
be offered with conﬁdence.

A general comment about the study, overall, should be made. Few variables
showed any relationship to ATP scale scores. The few correlations that were signiﬁcant,
were weak; no coefﬁcient was greater than .47 - for worldview inﬂuence on attitudes

toward poverty and the tax preference subscale variable. The sample was sufﬁciently

93

homogeneous that a number of variables did not have more than a few students who
differed in their responses from the majority of the respondents - knowing someone poor
(almost all did), political party (too few Republicans to analyze for differences), being a
recipient of most forms of government assistance (too few in various assistance
categories to analyze). A larger sample of social work graduate students, drawn from
multiple schools of social work in multiple areas of the state or across regions of the
country, may lead to samples with enough heterogeneity to permit analyses having
reasonable levels of statistical power.

A ﬁnal note about measurement is also offered. Eight variables were assessed
using a modiﬁed magnitude estimation measurement technique. As described in the
Methods section, respondents were directed to mark a continuum, represented by a solid
line, with an X at that place that best described each of the following: Social
class/economic status just prior to starting the MSW program and for family of origin;
amount of thought given to working with the poor; political orientation varying between
Liberal and Conservative; and amount of inﬂuence religion/spiritual tradition had on
decision to become a social worker. Signiﬁcant correlations were detected between four
of these variables and the ATP scale score.

This measurement approach appears to hold promise as a means of generating
variables with an interval/ratio level of measurement that can then be analyzed with the
stronger, parametric statistics (Pearson r, Student’s t, ANOVA). The eight variables
examined in this study using magnitude estimation are generally assessed by offering
respondents ordinal level, ranked answer choices that are most appropriately analyzed

with less powerful non-parametric statistics (Spearrnan rho, Mann-Whitney U, Kruskal-

94

 

Wallis K). This modiﬁed use of magnitude estimation validates the technique as a tool
for gathering self-reported ratings on background, attitudinal, and other variables
commonly assessed by social science researchers.

Implications for social work education

The study intent was to craft an improved measure of values related to poverty
and economically disadvantaged people. The study purposes were achieved; an
enhanced, valid, and reliable scale of Attitudes Toward Poverty was tested and
exploratory analyses of MSW student sample characteristics were conducted. The import
of the study lies in the heuristic value of this preliminary research.

A logic model for evaluating social work education for values outcomes

The logic model for social work education, detailed in Chapter 1, provides a
roadmap for ongoing evaluation of educational outcomes in Social Work. The survey
can serve educational evaluation by providing outcome measures. Selected course
evaluations could include measuring changes in student beliefs, attitudes, opinions, and
preferences. These variables can be assessed at the start and end of a course, the
beginning and ending of the foundation year, the specialization year, or the entire period
of graduate education.

The ATP survey and subscales provide one vehicle for advancing the profession’s
journey to evidence-based educational practice. The survey and its subscales fulﬁll
several assessment tasks. The logic model included both inputs and outputs. Inputs were
comprised of students, faculty, and courses, among others. The ATP can be used to
assess the value orientation of incoming students, current faculty, and course content with

respect to poverty attitudes. The ATP also can be used to create a standard based on

95

 

social work educator attitudes toward poverty and the poor. The measure can be used to
craft a standard, informed by the faculty, by the CSWE, or some other group of
professionals, against which student perspectives can be compared.

The survey provides a beginning point for assessing the inﬂuence of the
profession’s beliefs. The ATP also offers a means for contrasting social work
professionals’ attitudes with those held by the general public. Such comparisons will
help to locate the profession within the larger exosystem. The scales also may be helpful
in addressing social workers in mesosystems within the social environment. More
speciﬁcally, the ATP can help the profession examine the inﬂuence of practitioners’
attitudes. Ultimately social workers’ attitudes need to be assessed within the dynamic
systems that form at social intersections where social workers meet the populations we
help.

Study Limitations

As with all research, this study had a number of limitations. Social work values
were presumed to be a unitary concept. Using the NASW Code of Ethics as “the” source
of the profession’s values risks hiding the multi-faceted opinions of multiple social work
practitioners and educators. Relying on the current Code of Ethics also misrepresents the
profession’s history that resides in the collective, public mind. This collective memory
also is a valid indicator of the profession’s values and ought not be disregarded. This
limitation delineates the larger challenge of locating indicators or selecting valid
measures of the social profession’s values. Despite this issue the results have heuristic
value in that they help to reﬁne the variables and questions to investigate in future

research.

96

 

The nature of the sample and the sampling procedure were not probability-based.
Probability sampling is needed to increase conﬁdence in the generalizability of ﬁndings.
Sample selection was timed poorly. Students in the ﬁrst HBSE theory course were
enlisted to the study twelve weeks into a ﬁfteen-week semester. Students already had
been introduced in the theory course to poverty issues, poverty statistics, and policies and
theories related to poverty and the poor. Therefore the sample was not drawn from the
population of students that, normally, would not have had exposure to graduate-level
information regarding the study topic.

This difference in topic exposure was not uniform across the four course sections.
Two sections had the same instructor and the other two classes had two different
instructors. Two of the three instructors had extensive experience with teaching the
HBSE theory courses. One instructor was new to the school and new to the theory course
material. In addition, many students, although not all, were enrolled simultaneously in a
policy and social welfare course required during the ﬁrst year of the full-time MSW
program. These academic differences affected the type of exposure students received to
the course readings and to faculty feedback during class discussions and on written
assignments.

In addition to sample-related issues, study limitations can be discussed with
respect to at least three different aspects of measurement in this research project. First,
the ATP scale items are worded with an absolute or global language that applies each
item too broadly. Twenty-four of the 37 scale items should be reworded to diminish the

absolute quality of endorsing an item about all poor people.

97

Second, the rating scale also is suspect as it relates to an agreement continuum.
Responses can vary between strong agreement and strong disagreement. The middle of
the ﬁve-point scale is neutral on the item. However neutral is akin to no opinion. A
middle anchor for the agreement rating could more realistically be phrased as equally
agree and disagree. Respondent attitudes may be more accurately represented by a
response choice that recognizes the complex issues and multiple realities underlying each
item.

Third, measures also imposed limitations on the research. Although the use of
magnitude estimation was intended to elevate the level of measurement for the variables
evaluated with this technique, the nature of the question response items likely was new to
respondents. Magnitude estimation is encountered most commonly in medical settings in
which patients are asked to use a pain scale from 0 to 10 where 10 is “the most pain
imaginable” and 0 is “no pain at all”. Being asked to provide estimates of social class,
liberalism and conservatism, amount of thought, and amount of inﬂuence on decisions is
a quite different application of magnitude estimation. The import of a lack of familiarity
with a measurement technique is difficult to ascertain. An additional challenge to
magnitude estimation is the lack of deﬁnite positional anchors. The beginning and
ending of the lines were not tied visually to the relative anchor words. A respondent’s
interpretation of the anchor positions was not assessed. Interpretation of the eight
magnitude estimation questions was left to individual subjectivity. The potential for
compromised validity exists for this measurement technique. However, the potential
compromise is outweighed by the increased level of measurement and enhanced

variability made possible using magnitude estimation.

98

With respect to the demographic, life experience, and professional interest
variables, a number of problems exist. A modiﬁed magnitude estimation technique was
used for the questions about social class/economic status, amount of thought given to
professional work with the poor, political orientation and extent of inﬂuence of
religious/spiritual afﬁliation on career decision. The validity of using this technique to
measure self-reports of subjective states such as perceptions, inﬂuence, quantity of
thought, and identiﬁcation, is not documented in the literature. While published support
is not essential before using novel measurement techniques, better justiﬁcation should
have been offered. Future use of this technique should be accompanied by explicit
discussion and justiﬁcation, and/or comparative use of categorical measures of the same
concepts.

The sample also was not large enough to yield a substantial number of individuals
with poverty related life experiences. Relatedly the lack of ﬁndings with respect to the
research questions about recipients of government assistance programs was most likely
related to subgroup sizes. With the exception of ﬁnancial aid recipients the subgroup
sizes were not sufﬁcient to detect moderately sized differences between recipients and
non-recipients of the different government assistance programs. Student samples will
have to be bigger to yield enough recipients that the power of statistical testing is not
compromised by small n’s.

Directions for future research

Certainly the limitations mentioned above should be addressed in any future

investigation of social work student attitudes toward poverty and the poor. With respect

to the Atherton, et al. (1993) Attitudes Toward Poverty Scale, an elaboration of the scale

99

should focus on rephrasing items to diminish the need for respondents to endorse all
encompassing statements about all poor people. In addition, the response rating
categories should be modiﬁed so that the middle of the ﬁve response choices is changed
from neutral to equally agree and disagree or both agree and disagree. Items and
response choices that recognize the complexity of the issues underlying each item may
more accurately capture respondent attitudes.

Continued research also is needed to identify personal characteristics and
experiences of social work students and human service workers who hold more positive
attitudes toward the poor. These individuals may be disposed to serving disadvantaged
people more successfully and appropriately than individuals who hold less positive or
outright negative attitudes toward the poor. This information could be used to identify,
quite speciﬁcally, students and workers who have positive attitudes and offer poverty-
related ﬁeld training or employment opportunities. Students and workers with less
positive attitudes could be targeted for additional education or professional development.
Used in these ways, the Attitude Toward Poverty Scale could be helpful in identifying
and understanding the variability of attitudes among social workers and human service
providers. Future research should strive to meet the information needs and questions of
social work educators regarding student attitudes toward poverty and the poOr.

Despite its limitations the Attitude Toward Poverty scale shows promise as an
attitude assessment tool for social work educators, attitude researchers and evaluators.
Speciﬁcally, the scale can be used to evaluate the extent to which social work curricula
and courses meet the mandates of the Council on Social Work Education. The EPAS

Educational Polig for BSW and MSW programs states,

100

The purposes of social work are to enhance human well-being and alleviate

poverty, oppression, and other forms of social injustice. [and] to pursue

policies, services, and resources through advocacy and social or political actions

that promote social and economic justice (CSWE, 2003).
If the curriculum at any particular school of social work fulﬁlls the CSWE mandate for
educating students to have an understanding and appreciation for the plight and service
needs of disadvantaged individuals, then students’s attitudes reasonably could be
expected to become more positive during the time they are enrolled in a social work
program. Use of the Attitudes Toward Poverty scale for self-study or longitudinal
curriculum evaluation will help to inform faculty, academic administrators and CSWE
evaluate the effectiveness of professional training in meeting the Social Work
profession’s stated purpose and objectives for education.

The process of evaluating any program entails addressing the limits of change.
The restrictions that programmatic resources and processes impose on program functions
must be documented and recognized. The time has come for social work educators to
practice the techniques and meet the standards we teach our students to use. By avoiding
calls for accountability from within the communities we intend to serve, we very well
may be revealed, not as scientist-practitioners, but as a community of well-educated, high
minded, but shamefully misguided, professionals. The profession may ﬁnd, that we have
been marketing a product we do not or cannot produce, namely, alleviating poverty and
enhancing the well-being of the economically disadvantaged. Through meeting the
challenge of evidence-based practice, the profession will be able to claim that social work

does provide specialized knowledge in unique ways; ways that ultimately enhance human

well-being and diminish social injustice.

lOl

APPENDIX

102

APPENDIX A

Informed consent and study disclosure form
and

survey instrument

103

November 2004
Dear Social Work Student:

I am requesting your help with a research project. As a doctoral student in social work I am
currently working on my dissertation. I am conducting a study of social work student thoughts
about poverty and the poor. My project is a replication and expansion of research published in
1993 about student attitudes in the southern United States. The beneﬁts that will result from this
study do not accrue to you directly. This project helps me fulﬁll a doctoral program requirement.
The research also will provide ongoing information to the social work profession that can be used
to assess social work education about poverty and the poor. In the long run the results of this
study will be used to reﬁne a survey instrument that can be used to expand the knowledge base
for our profession.

Completing the questionnaire will take 20 minutes. Your decision to complete or not complete
the questionnaire will not affect your grade in this course either negatively or positively. Your
response to the questionnaire is conﬁdential; neither the instructor nor I will know which
questionnaire is yours because the questionnaire will not be identiﬁable as yours. Your
participation is completely voluntary and conﬁdential. Your identity and privacy will be
protected to the maximum extent allowable by law.

As with all research projects, there are risks and beneﬁts associated with this study. The risks to
you may involve the emergence of painful memories or strong emotions that people sometimes
experience when thinking about poverty and being poor. If you experience any distress or
troubling thoughts as a result of completing this survey I am available to you, as is your
instructor, to provide support. I can be reached at (517) 432-5912 or by e-mail sent to
peg.whalen@ssc.msu.edu .

If you have any questions about this study, please contact the investigator (Peg Whalen, 153
Baker Hall, MSU School of Social Work, East Lansing, 48824, (517) 432-5912, e-mail:
wg.whalen@ssc.msu.edu ). If you have questions or concerns regarding your rights as a study
participant, or are dissatisﬁed at any time with any aspect of this study, you may contact -
anonymously, if you wish —Peter Vasilenko, Ph.D., Chair of the University Committee on
Research Involving Human Subjects (UCRIHS) by phone: (517) 355-2180, fax: (517) 432-4503,
e-mail: ucrihs@msu.edu, or regular mail: 202 Olds Hall, East Lansing, MI 48824. The primary
investigator and supervising faculty member also is available to address your concerns. You may
contact Dr. Rena Harold, 254 Baker Hall, School of Social Work, MSU, East Lansing, MI
48824-1118, she also can be reached by phone at (517) 353-8616, or by email to
haroldr@msu.edu .

Your signature below indicates your voluntary agreement to participate in this study. This
informed consent form will be collected and stored separately from the questionnaire you
complete. Do not write your name on the questionnaire. Since each questionnaire is not

identiﬁable you will not be able to withdraw from the research once you have returned the
completed questionnaire.

Thank you in advance for your time.

Peg Whalen, MSW, ABD

104

51H! g l' .

Use the following scale to rate the extent of your agreement with the following statements.
Strongly Disagree (SD), Disagree (D), Neutral on the item (N), Agree (A), Strongly Agree (SA)

 

A person recervrng welfare should not have a nicer car than I SD I3 N A SA

2. Eigpeople will remain poor regardless of what’s done for SD D N A SA
3. Welfare makes people lazy. SD [3 N A SA
4. Any person can get ahead in this country. SD D N A SA
5. Poor people are satisﬁed receiving beneﬁts. SD 13 N A SA
6. 31125:? recipients should be able to spend their money as they SD D N A SA
7. ggtzbrlle-bodred person usrng food stamps rs nppmg off the SD [3 N A SA
8. Poor people are dishonest. SD D N A SA
9. If poor people worked harder, they could escape poverty. SD D N A SA
10. Most poor people are members of a minority group. SD D N A SA
1 1. People are poor due to circumstances beyond their control. SD 13 N A SA
12. Society has the responsibility to help poor people. SD D N A SA
13. People on welfare should be made to work for their beneﬁts. SD 13 N A SA
14. Unemployed poor people could ﬁnd jobs if they tried harder. SD D N A SA
15. Poor people are different from the rest of society. SD 13 N A SA
16. Being poor is a choice. SD D N A SA
17. Most poor people are satisﬁed with their standard of living. SD [3 N A SA
18. Poor people think they deserve to be supported. SD D N A SA
19. Welfare mothers have babies to get more money. SD 13 N A SA
20. Children raised on welfare will never amount to anything. SD D N A SA
21. Poor people act differently. SD D N A SA
22. Poor people are discriminated against. SD D N A SA
23. Most poor people are dirty. SD 13 N A SA
24. People who are poor should not be blamed for their SD D N A SA

misfortune.

105

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

If I were poor, I would accept welfare beneﬁts.

Out-of-work people ought to have to take the ﬁrst job that is
offered.

The government spends too much money on poverty
programs.

Some “poor” people live better than I do, considering all their
beneﬁts.

There is a lot of fraud among welfare recipients.

Beneﬁts for poor people consume a major part of the federal
budget.

Poor people use food stamps wisely.

Poor people generally have lower intelligence than non-poor
people.

Poor people should be more closely supervised.

I believe poor people have a different set of values than do
other people.

I believe poor people create their own difficulties.

I believe I could trust a poor person in my employ.

I would support a program that resulted in higher taxes to
support social programs for poor people.

Welfare discourages people from working.

Poor people give up looking for work too quickly.

Poor people are no different from me except in life
circumstances.

Poor people are more responsible for their condition than
many people would like to admit.

It is unfair that welfare mothers get more money if they have a
baby since working parents who have another child don’t
receive more money.

I believe people are poor because the US. economy excludes
them.

Poor people don’t care about their neighborhoods.

Poor people don’t keep up the properties they rent or own.

A person going on welfare should have to submit to urine or
blood tests for drugs.

I would feel uncomfortable visiting the home of a poor person.

I would support an increase in the sales tax to support social
programs for poor people.

I would support an increase in the state income tax to support
social program for poor people.

106

SD

SD

SD

SD

SD

SD

SD

SD

SD

SD

SD

SD

SD

SD

SD

SD

SD

SD

SD

SD

SD

SD

SD

SD

SD

2

Z Z Z Z Z Z Z

2 Z Z

2

SA

SA

SA

SA

SA

SA

SA

SA

SA

SA

SA

SA

SA

SA

SA

SA

SA

SA

SA

SA

SA

SA

SA

SA

SA

I would support an increase in the federal income tax to
support social program for poor people.

51. Many poor people do work but they get paid “under the table”. SD [3 N A SA

50. SDDNASA

The government should be able to limit the ways welfare and
food stamp recipients spend their beneﬁts.

Society should expect poor people to work in order to receive
beneﬁts.

More and better substance abuse treatment would result in

54' substantially less poverty. SD D N A SA

52. SDDNASA

53. SDDNASA

 

E 12! I '1' IE! I

 

 

For each of the following questions please choose the response that most clearly describes you.
If none of the answer choices listed are accurate feel free to write in your preferred response.

 

 

 

I. How old were you on your last birthday? years.

 

2. Please describe your race and write in your ethnicity or nationality.
(choose any / all that apply)

 

 

 

 

Cl African American / Black 0 American Indian
Cl European American / White D Chicano
CI Asian American/Paciﬁc Islander 0 Other Latino/Hispanic
CI lntemational student 0 Other (describe)
3. What gender are you? __ Female _ Male

4. On the lines below, mark with an X that place on the continuum which best
describes your socio-economic status or social class just prior to entering the
MS W program.

Prior to entering the MSW program I would describe myself as:

 

Poor / Lower Class Wealthy/ Upper Class

5. On the lines below, mark with an X that place on the continuum which best
describes the socio-economic status or social class of your family of origin when
you were growing up.

I would describe my family of origin when I was growing up as:

 

Poor / Lower Class Wealthy /Upper Class

 

107

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6. How do you describe your current spiritual / religious afﬁliation AND what was your
family’s spiritual / religious afﬁliation when you were growing up? (choose all that apply)
Denomination / Sect / , Denomination / Sect /
I am now . . . Type Iwas raised . . . Type
(please specify) (please specify)
D None CI None
Cl Catholic CI Catholic
0 Protestant U Protestant
D Other Christian El Other Christian
CI Jewish Cl Jewish
CI Muslim C] Muslim
Cl Other I D Other 1
Cl Other 2 CI Other 2
7. Are you registered to vote? _ Yes __ No
8. How regularly do you vote in US. elections?
CI Not Applicable CI November presidential elections
Cl November General Election each year CI Presidential primary elections
CI Primary election each year Cl Only particular local, county, city

CI All local, county, city elections each year elections
CI I never, or hardly ever, vote

9. What is your current political party afﬁliation? (circle only one)

None Democrat Independent Republican Other (describe)

 

10. On the line below, mark with an X the place on the continuum that most accurately
describes you

 

Far Left Liberal Conservative Far Right

I 1. Are you now, or have you ever been, poor? Don’t know No Yes

12. Would you say any of your parent(s) or grandparents(s) grew up poor?

Don’t Know No Yes
If yes, who?

 

I3. Have you ever personally known someone outside your family who was poor?

Don’t know No Yes

108

I4.

15.

l6.

l7.

18.

Have you had employment, internship, or volunteer related experience with poor people or

poverty?
Don’t know
No
Yes If yes, was your experience related to?

Employment Volunteer lntemship/Field Placement

Are you currently receiving or have you ever received any of the following:
(choose all that apply)

Cl Aid to Dependent Children (ADC/AFDC) CI Homeless Shelter

CI General Assistance / Welfare CI Unemployment Compensation
CI Food Stamps / Bridge Card CI Workers’ Compensation

0 Social Security Insurance (SSI) Cl Medicaid

D Social Security Disability Insurance (SSDI) Cl Housing Subsidy / Assistance
Cl Women, Infants & Children (WIC) Cl Veterans Beneﬁts

D Food Bank CI Federal / State / Private Student
CI Scholarship or Fellowship Grants / Loans (Pell, BEOG,
D Other (describe) Perkins, GSL, etc)

 

Are you currently, or do you plan to enroll, in a social work program, at MSU or
elsewhere, as a:

 

CI Clinical student, micro practice
CI OCP student, macro practice
D BSW student

0 Other (specify)

Cl Undecided

What professional interest do you have in working with poor people and/or with policies
or programs for poor people?
(circle the most accurate response choice or write in your answer)

I had not given it any thought.

No interest, I have a speciﬁc interest in other populations.

No interest, I would not / could not work with/for the poor.

Some interest, I would not mind working with/for the poor.

High interest, I speciﬁcally sought a social work degree in order to work with/for the
poon

Other

[3 DDUDU

 

 

On the line below, mark with an X the place on the continuum that indicates how much
thought you have given to working with poor people or with policies or programs for poor
people?

 

None A great deal

109

19.

20.

21.

22.

 

 

 

 

On the line below, mark with an X the place on the continuum that indicates to what
extent your religious/spiritual afﬁliation or heritage inﬂuence your attitudes about poverty
and the poor.
No inﬂuence at all Greatest inﬂuence of all
on my attitudes about poverty and the poor on my attitudes about poverty and the poor
On the line below, mark with an X the place on the continuum that indicates to what
extent your religious/spiritual afﬁliation or heritage inﬂuenced your decision to become a
social worker.
No inﬂuence at all Greatest inﬂuence of all
on my decision on my decision
On the line below, mark with an X the place on the continuum that indicates to what
extent your liberal or conservative views inﬂuence your attitudes about poverty and the
poor.
No inﬂuence at all Greatest inﬂuence of all
on my attitudes about poverty and the poor on my attitudes about poverty and the poor
On the line below, mark with an X the place on the continuum that indicates to what
extent your liberal or conservative views have inﬂuenced your decision to become a social
worker.
No inﬂuence at all Greatest inﬂuence of all
on my decision on my decision

 

- THANK YOU FOR PARTICIPATING -

110

 

REFERENCES
Addams, Jane. (1961). Twenty years at Hull House. NY: MacMillan.

Atherton, C., Gemmel, R. J. Haagenstad, S., Holt, D. J., Jensen, L.A., O’Haran, D. F. &
Rehner, T. A. (1993). Measuring attitudes toward poverty: A new scale. Social
Work Research & Abstracts, 29, 28-30.

Azjen, I. & Fishbein, M. (1980). Understanding attitudes and predicting social behavior.
Englewood Cliffs, NJ: Prentice-Hall.

Baggett, D.W. (1993). Study of the University of Massachusetts at Amherst faculty's
knowledge of disabilities, experience with educating students with disabilities,

and attitudes that faculty possess towards students with disabilities. Dissertation
Abstracts International, 56 (02), 452A. (U M1 No. 9518072)

Barlow, D.H., Hayes, S.C. & Nelson, R. O. (1984). The scientist practitioner: research
and accountability in clinical and educational settings. Boston: Allyn & Bacon

Beigi, G. (1998). Students' knowledge and attitudes toward public assistance and its
recipients. Dissertation Abstracts International, 36 (06),452A. (U M1 No.
1390094)

Bloom, M. & Fischer, J. (1982). Evaluating practice: Guidelines for the accountable
professional. Englewood Cliffs, NJ: Prentice-Hall.

Bloom, M., Fischer, J. & Orrne, J. (2003). Evaluating practice: Guidelines for the
accountable professional. New York: Allyn & Bacon.

Burghardt, S.& Fabricant, M. (1987). Working under the safety net : policy and practice
with the new American poor. Newbury Park: Sage Publications.

Campbell, D.T. (1963). Social attitudes and other acquired behavioral dispositions. In S.
Koch (Ed). Psychology: A study of a science (Vol. 6). New York: McGraw-Hill,
94-172.

Carrillo, D.F., Holzhalb, C.M. & Thyer, BA. (1993). Assessing social work students
attitudes related to cultural diversity: A review of selected measures. Journal of
Social Work Education. 29, 263-268.

Council on Social Work Education [CSWE]. (1998). Curriculum Policy Statement for
Master 's Degree Programs in Social Work Education. Alexandria, VA: Author.

Council on Social Work Education [CSWE]. (2003). Educational policy and academic
standards. Alexandria, VA: Author.

Cournoyer, B. (1991). The social work skills workbook. Belmont, CA: Wadsworth.

lll

Cournoyer, B. (2004). The social work skills workbook. (3rd ed.). Belmont, CA: Brooks-
Cole

Crosby, F., Pufall, A., Snyder, R., O'Connell, M. & Whalen, P. (1989). The denial of
personal disadvantage among you, me, and all the other ostriches. In M.
Crawford & M. Gentry (Eds.)., Gender and thought: Psychological perspectives.
NY: Springer-Verlag. 79-99.

Crosson-Tower, K. (2001). Exploring child welfare: A practice perspective. Boston:
Allyn & Bacon.

Cryns, AG. (1977). Social work education and student ideology: A multivariate study of
professional socialization. Journal of education for social work. 13, 44-51.

Day, P. J. (1997). A new history of social welfare (2“! Ed). Boston: Allyn & Bacon.
Dean, Jr., W. (1977). Back to activism. Social Work. 22, 369-373.
Doob, L.W. (1947). The behavior of attitudes. Psychological review. 54, 135-156.

Edwards, L. M. (1987). Biographies: Wald. In A. Minahan (Ed.). Encyclopedia of Social
Work. Silver Spring, MD: National Association of Social Workers.

Epstein, I. (1968). Social work and social action: Attitudes toward social action
strategies. Social Work, 13, 101-108.

Evans, S. M. (1997). Born for Liberty: A history of women in America. NY: Free Press.

Fields, B. J. (1987). Slavery, race, and ideology in the United States of America New
Left Review, 181, 95.

Fischer, J. (1983). Evaluation of social work effectiveness: Is positive evidence always
good evidence? Social Work 28, 74-77.

Fishbein, M. & Azjen, I. (1975). Belief attitude, intention and behavior: An introduction
to theory and research. Reading, MA: Addison-Wesley.

Gambrill, E. D. (1999). Evidence-based practice: An alternative to authority-based
practice. Families in Society, 80, 341-350.

Gambrill, E. D. (2000). Honest brokering of knowledge and ignorance. Journal of
Social Work Education, 36, 387-397.

Gambrill, E. D. (2001). Evaluating the quality of social work education: Options galore.
Journal of Social Work Education, 3 7, 418-429.

Gambrill, E. D. (2004). Evaluating the outcomes of social work practice: A pilot
program. Journal of Social Work Education, 40, 355-363.

112

Gans, H]. (1995). The war against the poor: The underclass and antipoverty policy.
New York: Basic Books.

Glicken, M. (2005). Improving the effectiveness of the helping professions: An evidence-
based approach to practice. Thousand Oaks, CA: Sage Publications.

Goffrnan, E. (1963). Stigma: Notes on the management of spoiled identity. Englewood
Cliffs, NJ: Prentice-Hall.

Grimm, L. G. & Orten, J. (1973). Student attitudes toward the poor. Social Work 18,
94-100.

Guttmann, D. & Cohen, B. (1992). Teaching about poverty in Israeli schools of social
work. International Social Work. 35 , 49-63.

Hayes, D.D. & Varley, B.K. (1965). Impact of social work education on student’s values.
Social Work 10, 40-46.

Holden, G., Meenaghan, T., Anastas, J. & Metrey, G. (2002). Outcomes of social work
education: The case for social work self-efﬁcacy. Journal of Social Work
Education, 38, 1 15-133.

IASWR (2001). IASWR’s three R’s. IASWResearch...reports....resources. 1, 1.

Jansson, B. (1988). The reluctant welfare state: A history of American social welfare
policies. Belmont, CA: Wadsworth.

Jansson, B. (2001). The reluctant welfare state: A history of American social welfare
policies (3rd ed.). Belmont, CA: Wadsworth.

Kaiser, HF. (1974). An index of factorial simplicity. Psychometrika, 39, 31-36.

Kapp, S. (2004). A logic model as an evaluation tool. Presentation at Michigan State
University, School of Social Work, E. Lansing, MI.

Kingﬁsher, C. (1996). Women in the American welfare trap. Phila, PA: University of
Pennsylvania Press.

Leiby, J. (1987). History of social welfare. In A. Minahan (Ed). Encyclopedia of Social
Work. Silver Spring, MD: National Association of Social Workers.

Leiby, J. (1978). A history of social welfare and social work in the United States.
NYzColumbia University Press.

Leighninger, L. (1987). Social work: Search for identity. New York: Greenwood Press.

Lerner, B. (1973). Democratic values and therapeutic efﬁcacy: A construct validity study.
Journal of Abnormal Psychology. 82, 491-498

113

Levine, R. A. (1964). Treatment in the home. Social Work, 9, 19-28.

Likert, R.A.(1932). A technique for the measurement of attitudes. Archives of
Psychology. 140, 5-53.

Lodge, M. (1981). Magnitude scaling, quantitative measurement of opinions. Beverly
Hills, CA: Sage Publications.

Loewenberg, F.M., Dolgoff, R., & Harrington, D. (2000). Ethical decisions for social
work practice. 6th Edition. Itasca, IL: F.E. Peacock Publishers.

Longres (1987). Biographies: Mary Ellen Richmond. In A. Minahan (Ed.).
Encyclopedia of Social Work. Silver Spring, MD: National Association of Social
Workers.

Lunbeck, E. (1994). The psychiatric persuasion: Knowledge, gender, and power in
modern America. Princeton, NJ: Princeton University Press.

Macarov, D. (1981). Social work students' attitudes toward poverty: a tri-national
study. Contemporary Social Work Education. 1, 150-160.

Margolin, L. (1997). Under the cover of kindness: The invention of social work
Charlottesville, VA: University Press of Virginia.

Melnick, J .A. (2000). The impact of curriculum on graduate social work students'
attitudes toward the poor. Dissertation Abstracts International, 61 (06), 452A.
(UMI No. 9976360).

Meyer, C. (1981). Social work purpose: status by choice or coercion? Social Work. 26,
69-75.

Miah, M. M. R. & Newcomb, P. R. (1995). Outcome measures in social work education
in the United States: a national survey. International Social Work, 38, 79-86.

Miller, J .B. (1976). Toward a new psychology of women. Boston: Beacon Press.

Mofﬁc, H., Brochstein, J ., Blattstein, A., & Adams, G. (1983). Attitudes in the provision
of public sector health and mental health care. Social Work in Health Care. 8,
17-28.

Morris, Robert. (1977). Caring for vs. caring about people. Social Work. 22, 353-359.

Mulroy, E.A. & Lauber, H. (2004). A user-friendly approach to program evaluation and
effective community interventions for families at risk of homelessness. Social
Work, 49, 573-586.

NASW. (2000). Code of Ethics. Washington, DC: National Association of Social
Workers Press.

114

National Association of Social Workers Center for Workforce Studies. (2005). Assuring

the sufficiency of a frontline worldorce: Preliminary report. Silver Springs, MD:
Author.

Neuman, W.L. & Kreuger, L.W. (2003). Social work research methods: Qualitative and
quantitative applications. Boston: Allyn & Bacon.

Orten, J. (1979). Experimentally inﬂuenced changes in student's attitudes toward the
poor. Ph.D. Dissertation, University of Alabama, May.

Orten, J. (1981). Inﬂuencing attitudes: A study of social work students. Social Work
Research & Abstracts. 1_7, 11-17.

Orten, J. (1976). Changes in students’ attitudes toward the poor. University of
Tennessee. [unpublished].

Peebles, J. (2000). The future of psychotherapy outcome research: Science or political
rhetoric? Journal of Psychology. 134, 659-670.

Piliavin, 1., Masters, S. & Corbett, T. (1979). Factors inﬂuencing errors in AFDC
payments. Social Work Research & Abstracts. 15, 3-17.

Pilsecker, C. (1978). Values: A problem for everyone. Social Work. 23, 54-57.

Piven, F.F. & Cloward, RA. (1993). Regulating the poor. (2"d ed.). New York: Vintage
Books.

Proctor, E. (1990). Evaluating clinical practice: Issues of purpose and design. Social
Work Research and Abstracts. 26, 32-40.

Quam, J .K. (1987). Biographies: Addams, Abbott, Reynolds. In A. Minahan (Ed.).
Encyclopedia of Social Work. Silver Spring, MD: National Association of Social
Workers.

Reeser, L. C. & Epstein, I. (1987). Social workers' attitudes toward poverty and social
action: 1968-1984. Social Service Review. 61, 610-622.

Reeser, L. C. (1986). Professionalization and social activism. Dissertation Abstracts
International, 47 (01), A. (UMI No. 8605939).

Rehner, T., Ishee, J ., Salloum, M. & Velasques, D. (1997). Mississippi social workers'
attitudes toward poverty and the poor. Journal of Social Work Education. 33,
131-142.

Reynolds, BC. (1963). An uncharted journey: Fifty years of growth in social work
New York: Citadel Press.

115

Roff, L., Adams, J. & Klemmack, D. (1984). Social work students' willingness to have
government help the poor. Arete. 9, 9-20.

Rokeach, M. (1968). Beliefs, attitudes and values: A theory of organization and change.
San Francisco: Jossey-Bass.

Rosenthal, B. (1993). Graduate social work students' beliefs about poverty and attitudes
toward the poor. Journal of Teaching in Social Work. 7, 107-121.

Rossi, P., Freeman, H. & Lipsey, M. (1999). Evaluation: A systematic approach, 3rd
edition Beverley Hills, CA: Sage Publications.

Rubin, A. & Babbie, E. (2001). Research methods for social work, 4th edition, Belmont,
CA: Wadsworth.

Santangelo, AC. (1992). The 1991 class of entering graduate students in California ’s ten
schools and departments of social work: a study of student attitudes toward the
poor, the social work ﬁeld, and characteristics influencing career decisions.
Unpublished doctoral dissertation, University of California at Berkeley.

Schwartz, S. & Robinson, M. (1991). Attitudes toward poverty during undergraduate
education. Journal of Social Work Education. 27, 290-296.

Sharwell, GR. (1974). Can values be taught? A study of two variables related to
orientation of social work graduate students toward public dependency. Journal
of Education for Social Work 10, 99-105.

Sherrod, D. (1982). Social Psychology. New York: Random House.

Sowell, T. (1987). A conﬂict of visions: Ideological origins of political struggles. New
York: Basic Books.

Specht, H. & Courtney, M. (1994). Unfaithful Angels. NY: The Free Press.

Stuart, P. H. (1990). Settlement houses: Changing sources of innovation in social work.
In H.H. Weisman (Ed.). Serious play. Silver Spring, MD: NASW.

Szasz, T. (1994). Cruel compassion: Psychiatric control of society ’s unwanted. New
York: John Wiley.

Tamoush, EM. (1998). Attitudes toward the poor: The impact of group versus individual
stimuli and participant demographics. Dissertation Abstracts International, 59
(04), 652B. (UMI No. 9831042).

Thurstone, LL. (1931). The measurement of attitudes. Journal of A bnormal and Social
Psychology, 26, 249-269

116

 

Thyer, B. A., Isaac, A., & Larkin, R. (1997). Integrating research and practice. In M.
Reisch & E. Gambrill (Eds). Social work practice in the 21st century (pp. 311-
316). Thousand Oaks, CA: Pine Forge Press.

Trattner, W. I. (1994). From poor law to welfare state: A history of social welfare in
America, 5'h Edition New York: Free Press.

Trattner, W. I. (1984). From poor law to welfare state: A history of social welfare in
America. NY: Free Press.

Tropman, J. E. (1989). American values and social welfare: Cultural contradictions in
the welfare state. Englewood Cliffs, NJ: Prentice Hall.

Tropman, J. E. (1995). The Catholic ethic in American society: An exploration of
values. San Francisco: Jossey-Bass Publishers.

Varley, B.K. (1963). Social work values: Changes in value commitments of students

from admission to MSW graduation. Journal of Education for Social Work 4,
67-76.

Walkowitz, DJ. (1999). Working with class: Social workers and the politics of middle-
class identity. Chapel Hill, NC: University of North Carolina Press.

Watson, F. D. (1922). The charity organization movement in the United States: A study in
American philanthropy. New York: NY, The Macmillan Company.

Wyers, N. (1978). Attitudes of income maintenance personnel: Old paradigms
revisited. California Sociologist. I, 151-167.

Zerhouni, EA. (2003). NIH Plan for Social Work Research. Washington, DC: NIH,
Department of Health and Human Services

117

   

iijiiiiijjitrjiiliijjii