Adolescence, emotional and behavioral problems, and counseling

MISSING IMAGE

Material Information

Title:
Adolescence, emotional and behavioral problems, and counseling where's the stigma?
Physical Description:
xi, 117 leaves : ill. ; 29 cm.
Language:
English
Creator:
Danda, Caroline Elder
Publication Date:

Subjects

Subjects / Keywords:
School Health Services   ( mesh )
Counseling -- Adolescent   ( mesh )
Adolescent Behavior   ( mesh )
Attitude to Health -- Adolescent   ( mesh )
Emotions -- Adolescent   ( mesh )
Mental Disorders -- Adolescent   ( mesh )
Department of Clinical and Health Psychology thesis Ph.D   ( mesh )
Dissertations, Academic -- College of Health Professions -- Department of Clinical and Health Psychology -- UF   ( mesh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D)--University of Florida, 2002.
Bibliography:
Bibliography: leaves 109-116.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Caroline Elder Danda.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 026732001
oclc - 53949367
System ID:
AA00022637:00001

Table of Contents
    Title Page
        Page i
    Acknowledgement
        Page ii
    Table of Contents
        Page iii
        Page iv
    List of Tables
        Page v
        Page vi
    List of Figures
        Page vii
    Abstract
        Page viii
        Page ix
    Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
    Conceptual bases for the current study
        Page 19
        Page 20
        Page 21
        Page 22
    Purpose of the study
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
    Method
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
    Analysis plan
        Page 42
        Page 43
        Page 44
    Results
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
    Discussion
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
    Appendix A. Vignette descriptions
        Page 97
        Page 98
    Appendix B. Adjective checklist
        Page 99
    Appendix C. Original activity preference list
        Page 100
    Appendix D. Adolescent activity preference list
        Page 101
    Appendix E. Manipulation check questions
        Page 102
    Appendix F. Help seeking scale
        Page 103
        Page 104
        Page 105
    Appendix G. Letter to parents
        Page 106
    Appendix H. Cover letter to adolescent
        Page 107
    Appendix I. Debriefing statement
        Page 108
    List of references
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
    Biographical sketch
        Page 117
        Page 118
        Page 119
        Page 120
Full Text











ADOLESCENCE, EMOTIONAL AND BEHAVIORAL PROBLEMS, AND
COUNSELING: WHERE'S THE STIGMA?
















By

CAROLINE ELDER DANDA


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2002















ACKNOWLEDGMENTS


There are many people who contributed to the successful completion of this

project. I would like to thank my committee members, with special thanks to my cochair

Suzanne Bennett Johnson for her enthusiasm and insight and for being an excellent role

model. Special recognition goes to my chair Garret Evans, whose encouragement,

thoughtful guidance, and confidence in me throughout graduate school and this project

have been instrumental in developing my identity as a researcher and clinician.

Data collection would not have been possible without the cooperation of Gloria

Spivey as well as the staff, teachers, and students at Columbia High School. Additional

recognition goes to the team of people who helped with data collection and/or data entry,

including Amy Baughcum, Kim Kirkpacktrick Justice, Rebecca Kristal, Rachel Nitzberg,

Abby Sia, Jessica Stilley, and Brenda Wiens.

I would also like to express my gratitude to my good friend and colleague, Beth

Bryant. Together, we weathered many storms and celebrated many milestones. Thanks

also go to my parents, who told me I could achieve anything that I wanted if I worked

hard enough and supported my decision to enter the field of psychology. Last, but most

certainly not least, I thank my husband and best friend, Matt, who unconditionally

supported and encouraged me through the stress of the dissertation and graduate school

and was willing to make the sacrifices that are an inherent part of being a spouse of

someone going through these processes.















TABLE OF CONTENTS


page

ACKNOW LEDGM ENTS................................................................................................... ii

LIST OF TABLES .............................................................................................................. v

LIST OF FIGURES...........................................................................................................vii

ABSTRACT .....................................................................................................................viii

INTRODUCTION ............................................................................................................... 1

Prevalence of M ental Disorder in Children..................................................................... 1
Service Utilization........................................................................................................... 2
Attitudes Toward M ental Illness ..................................................................................... 8
Attitudes Toward Counseling/Therapy......................................................................... 12

CONCEPTUAL BASES FOR THE CURRENT STUDY............................................... 19

Stigma ............................................................................................................................ 19
Attitudes ........................................................................................................................ 20
Vignette M ethodology................................................................................................... 21

PURPOSE OF THE STUD Y ............................................................................................ 23

Research Questions........................................................................................................ 23
Hypotheses..................................................................................................................... 24

M ETHOD ..........................................................................................................................31
Experimental Conditions............................................................................................... 31
Procedure....................................................................................................................... 37
Participants .................................................................................................................... 39

ANALYSIS PLAN ............................................................................................................ 42






iii









R E SU L T S ..........................................................................................................................45

D ata Inspection ....................................................................... ....................................... 45
Prim ary ANOVA Analyses...........................................................................................49
Secondary ANOVA Analyses....................................................................................... 56

D ISC U SSIO N ................................................................................................................... 75

APPENDIX A VIGNETTE DESCRIPTIONS............................................................ 97

APPENDIX B ADJECTIVE CHECKLIST................................................................ 99

APPENDIX C ORIGINAL ACTIVITY PREFENCE LIST..................................... 100

APPENDIX D ADOLESCENT ACTIVITY PREFERENCE LIST......................... 101

APPENDIX E MANIPULATION CHECK QUESTIONS ...................................... 102

APPENDIX F HELP SEEKING SCALE................................................................. 103

APPENDIX G LETTER TO PARENTS................................................................... 106

APPENDIX H COVER LETTER TO ADOLESCENT............................................ 107

APPENDIX I DEBRIEFING STATEMENT.......................................................... 108

LIST OF REFERENCES ................................................................................................ 109

BIOGRAPHICAL SKETCH........................................................................................... 117





















iv















LIST OF TABLES


Table Page

1. Demographic Characteristics of Participants (N=442) ............................................. 41

2. Subject Distribution across Experimental Conditions .............................................. 41

3. Means and Standard Deviations across Experimental Conditions on the Adjective
Checklist (N =434)............................................................................................... 63

4. Means and Standard Deviations across Experimental Conditions on the
Adolescent Activity Preference List (N = 433)................................................... 63

5. Respondent Characteristics of Participants Across Gender (N = 442) ..................... 64

6. Means on the Adjective Checklist and Adolescent Activity Preference List for
Emotional and Behavioral Problems Main Effect............................................... 64

7. Estimated Marginal Means on the Adjective Checklist for Emotional and
Behavioral Problems x Counseling Status Interaction........................................ 66

8. Estimated Marginal Means on the Adjective Checklist for Emotional and
Behavioral Problems x Gender Interaction......................................................... 67

9. Estimated Marginal Means on the Adolescent Activity Preference List for
Emotional and Behavioral Problems x Gender x Prior Counseling Experience
Interaction............................................................................................................ 69

10. Estimated Marginal Means on the Adjective Checklist for Emotional and
Behavioral Problems x Counseling Status x Perceived Similarity Interaction... 70

11. Estimated Marginal Means on the Adolescent Activity Preference List for
Emotional and Behavioral Problems x Gender x Perceived Similarity
Interaction............................................................................................................ 71

12. Estimated Marginal Means on the Adolescent Activity Preference List for
Ethnicity x Emotional and Behavioral Problems Interaction.............................. 72









13. Estimated Marginal Means on the Adjective Checklist for Counseling Status x
Gender x Grade Interaction................................................................................. 73

14. Estimated Marginal Means on the Adjective Checklist for Rurality x Counseling
Status Interaction................................................................................................. 74















LIST OF FIGURES


Figure Page

1. Main Effect of Emotional and Behavioral Problems on the Adjective Checklist..... 65

2. Main Effect for Emotional and Behavioral Problems on the Adolescent Activity
Preference List..................................................................................................... 65

3. Interaction Effect for Emotional and Behavioral Problems x Counseling Status on
the Adjective Checklist ....................................................................................... 66

4. Interaction Effect for Emotional and Behavioral Problems x Gender on the
A djective Checklist ............................................................................................. 67

5. Interaction Effect for Emotional and Behavioral Problems x Counseling Status x
Gender on the Adjective Checklist (see Table 3 for Means).............................. 68

6. Interaction Effect for Emotional and Behavioral Problems x Prior Counseling
Experience x Gender on the Adolescent Activity Preference List...................... 69

7. Interaction Effect for Perceived Similarity x Emotional and Behavioral Problems
x Gender on the Adjective Checklist................................................................... 70

8. Interaction Effect for Perceived Similarity x Emotional and Behavioral Problems
x Gender on the Adolescent Activity Preference List......................................... 71

9. Interaction Effect for Ethnicity x Emotional and Behavioral Problems on the
Adolescent Activity Preference List................................................................... 72

10. Interaction Effect for Rurality x Counseling Status on the Adjective Checklist...... 74















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

ADOLESCENCE, EMOTIONAL AND BEHAVIORAL PROBLEMS, AND
COUNSELING: WHERE'S THE STIGMA?


By

Caroline Elder Danda

August 2002

Chair: Garret D. Evans
Cochair: Suzanne Bennett Johnson
Major Department: Clinical and Health Psychology

Using a vignette methodology, this study examined adolescent attitudes toward

common emotional and behavioral problems and school-based counseling within a rural

community. Participants included 442 high school students from rural Florida. Peers

described as exhibiting behaviors associated with depression, family conflict, and

conduct disorder were consistently rated less favorably than peers without any problems,

with particularly strong effects for conduct disorder. Stigma (i.e., negative attitudes)

attached to emotional and behavioral problems appeared stronger than the stigma

attached to counseling. Although peers without apparent psychological problems were

rated less favorably if they attended counseling, they were still rated significantly more

favorably than peers with problems, regardless of their counseling status. Gender,

rurality, perceived similarity, prior counseling experience, and ethnicity moderated some

of the effects related to emotional and behavioral problems and counseling status. For









example, adolescents characterizing themselves as living in the country rated peers who

attended counseling less favorably than those living in town, which suggests that

adolescents living in the country may be less tolerant or understanding of peers who

attend counseling. Males and females also had significantly stronger and opposite

reactions to their same-sex peers with depression. Females viewed their same-sex peers

with depression in counseling most positively, and males viewed their same-sex peers

with depression less positively compared to those with family conflict but not conduct

disorder.

This study indicated that the manifestations of emotional and behavioral problems

seem to impact attitudes more than counseling, per se. Consequently, the message that

counseling can help ameliorate emotional and behavioral problems that might already be

negatively impacting social relationships, and seems to have little additional negative

effects with respect to adolescent attitudes, can help allay adolescent and parental

concerns about their children attending counseling with a psychologist. Education about

the nature and prevalence of common emotional and behavioral problems as well as the

effectiveness of counseling in addressing these problems should be an integral part of any

school-based mental health program in order to foster more open discussion of mental

health issues, and perhaps lead to less stigmatization and more support among the peer

group.















INTRODUCTION

Prevalence of Mental Disorder in Children

Estimates of adolescent emotional and behavioral difficulties range from 11-22%

(Cohen & Hesselbart, 1993; Kashani et al., 1987; Offord et al., 1987; Whitaker et al.,

1990). Emotional disturbances begin to occur at increased rates in adolescence

(Achenbach, Howell, Quay, & Conners, 1991; Dubow, Lovko, & Kausch, 1990; Offer &

Schonert Reichl, 1992), with depression, eating disorders, family conflict, conduct

disorder, and substance abuse among the most frequently reported problems (Anglin,

Naylor, & Kaplan, 1996; Dubow et al., 1990; Kashani et al., 1987; Whitaker et al., 1990).

Estimates from the Office of Technology Assessment (1991) further suggest that 18-22%

of adolescents have emotional disorders that warrant intervention.

With respect to gender, overall prevalence rates of emotional and behavioral

problems are relatively equal. Although males tend to have higher prevalence rates of

disorder in childhood, females evidence increased symptomatology and distress (e.g.,

emotional disorders) during adolescence, leading to comparable prevalence rates between

the sexes during the adolescent years (Dubow et al., 1990; Offord et al., 1987). However,

male and female prevalence rates differ by type of disorder. Adolescent males tend to

exhibit more externalizing difficulties (e.g., behavior disorders and substance abuse)

whereas adolescent females exhibit more internalizing difficulties (e.g., eating disorders,

dysthymia, and depression) (Ostrov, Offer, & Howard, 1989; Whitaker et al., 1990).









Compared to urban areas, prevalence rates of child and adolescent mental health

problems in rural areas do not differ greatly (Farmer, Stangi, Bums, Costello, & Angold,

1999), despite living conditions being characterized as idyllic in rural settings. In

addition, declines in traditional economic bases (e.g., farming, resource extraction, and

low-technology manufacturing) and related factors, such as poverty, substandard housing,

and family disruption, have placed some rural residents at higher risk for emotional

disorders (Kelleher, Taylor, & Ricker, 1992; Kenkel, 1986). Indeed, the farm crisis in the

1980s and mid-1990s and the associated rural economic decline have been associated

with increases in withdrawal, depression, family violence, and substance use among rural

adults and adolescents (Garfinkel, Hoberman, Pasons, & Walker, 1988, as cited in

Kelleher et al., 1992; Wagenfeld, Murray, Mohatt, & DeBruyn, 1994).


Service Utilization

Despite high prevalence rates of emotional and behavioral problems among

adolescents, the majority of these youth receive either inadequate or no mental health

services at all (Leaf et al., 1996; Offord et al., 1987; Whitaker et al., 1990). Only about

one-third to one-half of youths who need services actually receive them (Surgeon

General, 2000; Office of Technology Assessment, 1991). For example, in a community

study of self-reported mental health needs and help-seeking, only half of those who

indicated a need for help actually sought it (Saunders, Resnick, Hoberman, & Blum,

1994). Similarly, Rickwood and Braithwaite (1994) reported that approximately one-

quarter of adolescents identified themselves as distressed on a self-report measure, but

only 17% of these individuals sought professional help. Although self-report

methodologies and degree of impairment may inflate rates of emotional and behavioral









problems, studies using structured interviews and diagnostic criteria for mental disorders

have found similarly low levels of utilization. Whitaker and colleagues (1990) found that

less than half of adolescents who screened positive for a lifetime diagnosis of a disorder

and were currently experiencing impairments in functioning had received any

professional services. Likewise, Offord and colleagues (1987) found that only 16 % of

children with a diagnosable disorder had received services within the last 6 months.

Studies of service utilization, however, often employ broadly inclusive definitions

of mental health professionals and mental health contacts. For example, Rickwood and

Braithwaite (1994) included family doctor, general mental health services, and

educational help services in the category of professional help. Others have not specified

the source of help (Saunders et al., 1994) or the nature of the help; in such cases, a

reported mental health contact may not be a valid indicator of actual treatment received

(Leaf et al., 1996). The lack of definition related to mental health service use likely leads

to overestimation of service use. For example, Zahner, Pawelkiewicz, DeFrancesco, and

Adnopoz, (1992) found that although 51% of children who were identified as having

emotional and behavioral problems by both teachers and parents reported using services

for such problems, only 11% used services within the mental health sector. Offer,

Howard, Schonert, and Ostrov, (1991) found that among 500 adolescents, 22% reported

experiencing emotional and behavioral difficulties, but only one-third of these

adolescents sought help from a mental health provider, and only 20% actually entered

treatment (i.e., participated in more than 3 sessions).

Finally, some evidence suggests that youths with severe and/or multiple disorders

are more likely to receive services, but as many as two-thirds still remain untreated (Leaf









et al., 1996). Moreover, if most mental health treatment has been directed toward those

with more chronic, severe mental disorders, an even greater gap in service delivery exists

for children and adolescents with lower levels of symptomatology who still demonstrate

significant impairment (e.g., poor academic performance, problematic interpersonal

relationships, etc.). Essentially, these are the youths who are at risk for developing more

serious, chronic emotional and behavioral disorders that could be prevented if they

received early treatment. Indeed, the 1991 "Adolescent Health. Volume I. Summary and

Policy Options" report (Office of Technology Assessment, 1991) suggested that

adolescents who have subclinical mental health problems and who are unaware of

available services are at the greatest risk of falling through the cracks. Given that more

than two-thirds of adolescents do not receive sufficient or adequate treatment, greater

efforts need to be directed toward providing mental health services to these youths and

encouraging the use of these services among adolescents.


School-Based and School-Linked Mental Health Services

Schools often serve as the first point of contact for mental health problems and

subsequent referrals. Several researchers have even suggested that schools have become

the "de facto" mental health system, despite the fact that they may have little overlap with

specialty mental health services (Bums et al., 1995; Leaf et al., 1996; Zahner et al.,

1992). Integrated systems of care within the school setting, such as school-based and

school-linked health services, initially targeted physical health concerns; however, mental

health services were gradually incorporated, in part due to increased attention to

adolescent problem areas, such as teen pregnancy, risk-taking behaviors, substance abuse,

violence, depression, and suicide (Flaherty, Weist, & Warner, 1996). Partnerships and









interagency collaborations between schools and community/academic mental health

providers have allowed for the inclusion of mental health services in school-based and

school-linked clinics.

Currently over 900 school-based and school-linked health centers operate

nationwide, the majority of whom (86%) provide some type of counseling services

(Fothergill & Ballard, 1998). School-based and school-linked services may be

particularly well-suited for rural areas, which typically have fewer mental health

resources and more fragmented service delivery systems than urban and suburban

communities (Cohen & Hesselbart, 1993; Kelleher et al., 1992). Schools offer centralized

locations that are ideal for initiating referrals, coordinating a system of care, providing

continuity of care, and promoting better treatment planning and generalization. Therapists

can speak directly with guidance counselors or teachers to gain more information about

the nature of a problem a child is experiencing as well as collaborate with them to

facilitate generalization of treatment gains across settings. Students may also be more

amenable to counseling if suggested by guidance counselors with whom they have

frequent contact and a trusted relationship. Issues impeding follow-through with mental

health service referrals, such as transportation and poor communication with referral

sources, are far less problematic within a school-based system of care (Santelli, Morreale,

Wigton, & Grason, 1996). Cost also becomes less of a barrier as agencies often partner

with schools to provide mental health services at reduced fees or free of charge (G. D.

Evans & Rey, 2001; Fothergill & Ballard, 1998). Services that occur within the school

setting and/or during school hours also reduce the burden on the family to bring the child

to weekly appointments (Santelli et al., 1996). Moreover, mental health services in the









context of the schools provide a sense of familiarity and may eventually be considered

part of the routine services offered, thereby reducing the stigma that may be associated

with attending a mental health treatment center.

Data regarding service utilization suggest that school-based and school-linked

services represent a viable treatment option for adolescent mental health problems.

Research examining school-based health services (both physical and mental health)

reported that approximately 20-25% of all visits were related to mental health (Anglin et

al., 1996; Balassone, Bell, & Peterfreund, 1991). Some data even suggest increased

utilization of school-based mental health services relative to other forms of service

delivery, such as hospitals or outpatient clinics (Anglin et al., 1996; S. W. Evans, 1999).

Moreover, many adolescents perceived school-based mental health services as helpful,

approachable, and even confidential (Balassone et al., 1991; S. W. Evans, 1999), and a

recent outcome study demonstrated that youths who received school-based services

demonstrated decreased levels of depression and improvements in their self-concept

(Weist, Paskewitz, Warner, & Flaherty, 1996).


Barriers to Help-Seeking

Although school-based mental health services appear to be effective, a large

number of adolescents do not receive the treatment they need, often due to a reluctance to

seek help (i.e., counseling). Adolescents may be reluctant to seek services because of the

personal nature of problems as well as concerns about what goes on in therapy,

confidentiality, privacy, and feelings of embarrassment (Adelman & Taylor, 1993;

Balassone et al., 1991; Taylor, Adelman, & Kaser Boyd, 1985). The school setting can

certainly exacerbate these concerns about confidentiality and privacy. An adolescent may









be called out of class, or at least be absent from class, to attend counseling with a

psychologist. Peers may also have increased opportunities to see an individual enter the

therapist's office. Indeed, recent evidence suggests that confidentiality and privacy are of

utmost concern in the school setting. Riggs and Cheng (1988) found that adolescents

experiencing emotional problems would only use a school-based clinic if confidentiality

was guaranteed. Similarly, families unfamiliar with school-based services rated

confidentiality as a significant concern as compared to traditional clinic-based services

(S.W. Evans, 1996, as cited in S.W. Evans, 1999).

Confidentiality concerns are only heightened by the potential stigma associated

with receiving mental health counseling. Adolescents may be worried peer reactions to

the discovery that the adolescent is seeing a psychologist for counseling. Although Day

and Reznikoff (1980) reported that young boys (aged 7-12 years) referred for therapy

were concerned about neighbors or teachers finding out, scant attention has been given to

adolescents' attitudes toward mental health problems and counseling, and even less

attention to whether children and adolescents negatively evaluate or behave negatively

toward (i.e., stigmatize) individuals who have emotional and behavioral problems or

attend counseling.

Understanding adolescent attitudes and actions toward common emotional and

behavioral problems and counseling may be even more important in rural areas where

fears of being labeled as mentally ill or stigmatized have often been cited as a significant

barrier to seeking psychological treatment (Berry & Davis, 1978; Kenkel, 1986; Surgeon

General, 2000). Traditional rural values, such as self-reliance, autonomy, conservatism,

and distrust of outsiders are often incompatible with help-seeking (Hoyt, Conger, Valde,









& Weihs, 1997; Kelleher et al., 1992; Wagenfeld et al., 1994). Rural residents often rely

on informal sources for help (e.g., family, friends, religious leaders), ignore, deny, and

minimize problems (Kelleher et al., 1992; Kenkel, 1986), and tend to be less familiar

with routine therapy services. Prior to the establishment of Community Mental Health

Centers (CHMC) in 1963, the mental health needs of rural residents were virtually

ignored (Hargrove & Breazeale, 1993). The resources provided by CMHC, however,

were largely dedicated to treatment of the chronically and severely mentally ill. Given

that many rural residents' experience and knowledge of mental health treatment likely

stems from their perceptions of CHMCs, it is not surprising that they might equate all

mental health services with treatment for chronic and severe mental illnesses. Stigma

related to such mental health problems may infect stigma for all mental health services.

In this respect, many rural residents lack knowledge about existing mental health

services, common emotional and behavioral problems, and the effectiveness of

psychological counseling in addressing these problems (Berry & Davis, 1978).

The movement toward integrated school-based mental health services ameliorates

issues of accessibility and availability, particularly in rural areas, and raises the question

about barriers, such as stigma, that may impede adolescent utilization such services.


Attitudes toward Mental Illness

Adult Perceptions

Negative evaluations toward the severely and chronically mentally ill have been

widely documented among adults. Nunnally's (1961) classic studies found that the public

generally held negative views of the mentally ill, frequently ascribing to them

characteristics such as worthless, dirty, dangerous, cold, unpredictable, and insincere.









Early studies suggested that those with mental illness were to be feared, avoided, perhaps

controlled but also treated with kindness (Hayward & Bright, 1997). Even now, the

public often seems to have a restricted view of mental illness, limited to severe mental

illness, such as schizophrenia and other conditions requiring hospitalizations. Indeed,

Hayward and Bright (1997) argued that people are reluctant to label all but the most

extreme forms of behavior as mental illness, suggesting that the general population has a

predisposition to interpreting mental illness as a severe, chronic condition. Consequently,

it is not surprising that common public misconceptions include views of the mentally ill

as "homicidal maniacs who need to be feared," people with "childlike perceptions of the

world that should be marveled," and "rebellious, free spirits" (Corrigan & Penn, 1999, p.

766). Although recent conceptions of mental illness have expanded to include anxiety

and depression, many people still associate the term mental illness with a psychotic

condition that includes violent and unpredictable behaviors that should be feared

(Surgeon General, 2000).

Such negative evaluations, or stereotypes, may prevent people from seeking

treatment, particularly for less chronic and severe conditions, for fear of being labeled

and stigmatized as mentally ill (Hayward & Bright, 1997). Although stigma is often cited

as a barrier to seeking treatment, little research has examined stigma and attitudes

regarding individuals who seek outpatient treatment, particularly for less severe

behavioral and emotional problems. Although evidence suggests that current

psychotherapy clients may not perceive stigma associated with seeking help as a salient

problem (Halgin & Weaver, 1986), Stefl and Prosperi (1985) found that those in need of

services who did not utilize them anticipated levels of stigma (i.e., feared being looked









down upon or their neighbors and friends finding out) almost two times higher than

individuals who needed help and sought it. Thus, it appears that concerns about stigma

represent a significant barrier to help-seeking among individuals in need of mental health

services.


Child and Adolescent Perceptions

Studies of children's views of mental illness also suggest generally unfavorable

views. Compared to medically ill, handicapped, or learning-disabled youths, children rate

emotionally disturbed peers less favorably (Parish, Ohlsen, & Parish, 1978), view them as

less attractive (Novak, 1974; Roberts, Johnson, & Beidleman, 1984) and perceive such

peers as less similar to themselves (Novak, 1974). Even when children hold generally

positive perceptions of peers with mental disorders, they still attribute more negative

characteristics to them compared to children without a mental disorder (Friedrich,

Morgan, & Devine, 1996). Likewise Lopez (1991) found that while adolescents may not

necessarily see people with mental illnesses as dangerous, they still view them as

relatively unpredictable.

More detailed examinations of attitudes toward psychological problems suggest

attitudes may also vary according to type of problem. Children as young as elementary

school can differentiate between normal and deviant peers and often make distinctions

among disorders (Chassin & Coughlin, 1983; Coie & Pennington, 1976; Marsden, Kalter,

Plunkett, & Barr-Grossman, 1977). In particular, aggressive behavior is consistently rated

as more deviant and viewed more negatively (Coie & Pennington, 1976; Hoffman,

Marsden, & Kalter, 1977; Marsden et al., 1977). The greater negative perceptions of

aggressive behavior might be partially explained by a belief that aggressive behavior









represents a lack of self-control and adherence to social norms (Roberts et al., 1984) and

is consistent with research that has found aggression to be one of the strongest predictors

of rejection by the peer group (Coie, Dodge, & Kupersmidt, 1990). Further, research has

demonstrated that behaviors such as fighting, group disruption, and irritable tempers are

consistently related to negative peer evaluations (Inderbitzen-Pisaruk & Foster, 1990).

Shyness and withdrawn behavior, which could be related to depression, have also been

associated with children being neglected, but not rejected (Inderbitzen-Pisaruk & Foster,

1990).

With increasing age, conceptions of mental illness become more complex and

abstract. Older children make better distinctions between deviant behavior associated

with mental disorder versus behavior associated with social norm violations, demonstrate

greater insight about the causes of deviant behavior, and show greater differentiation

between types of mental disorders (Chassin & Coughlin, 1983; Coie & Pennington, 1976;

Offer & Schonert Reichl, 1992; Sigelman & Mansfield, 1992). For example, Sigelman

and Mansfield (1992) found that adolescents were less willing to see a psychologist for

less serious problems; reluctance to seek treatment was most evident when symptoms

were not psychological in nature or reflected nonconformity or normality. Younger

children consistently rated mental disorders more negatively in general, while adolescents

were less consistent in their ratings, suggesting that adolescents were better able to

differentiate among the disorders (Sigelman & Mansfield, 1992). Once children reach

adolescence, however, there may be fewer changes in their conceptions of mental

disorders (Lopez, 1991).









With respect to gender, adolescent females tend to have more favorable views

toward individuals with mental illnesses than males. For example, Lopez (1991) reported

that adolescent females viewed individuals with mental illness as more similar to

themselves, saw them as less of a threat to society, endorsed more nurturing attitudes

toward them, and were more willing to interact with them. Likewise, Norman and Malla

(1983) found that females demonstrated greater social acceptance of peers with

psychological problems as well as a stronger belief in psychosocial etiology and a more

positive prognosis for these problems.

In sum, both adults and children appear to hold somewhat negative views of

mental disorders. However, much of the research has focused on attitudes toward severe,

chronic mental disorders, such as schizophrenia, or toward individuals who have been

hospitalized for mental illness, rather than more common emotional and behavioral

problems. Research on children's conceptions of psychological disorders has been

conducted primarily with younger populations; much less attention has been directed

toward adolescent perceptions. Research in this area also tends to be qualitative and

descriptive, often assessing general knowledge of psychology rather than specific

information about common emotional and behavioral problems, or uses dated, adult

measures or adult vignettes. Even less is known about general attitudes toward therapy or

mental health counseling, per se.


Attitudes Toward Counseling/Therapy

Adult Perceptions

The literature on adult perceptions toward individuals who seek psychological

treatment suggests neither entirely positive nor negative attitudes. Crisp, Gelder, Rix,









Meltzer, and Olwen (2000) found that the public held relatively accurate opinions about

the ability to treat several mental health problems and held somewhat optimistic views

about their prognosis. Nunnally (1961) also found that the general public holds

moderately favorable attitudes toward mental health professionals, which suggests that

the treatment they offer may also be viewed positively. Evidence also exists suggesting

that individuals with emotional and behavioral problems who seek treatment may be

viewed as more competent (Dovidio, Fishbane, & Sibicky, 1985). Likewise, Schwarzer

and Weiner (1991) found that individuals with medical and psychological problems

received less blame or anger and more pity if they were perceived as actively coping with

their illness. Although examples of active coping were only provided for physical illness

(e.g., adherence to medical regimen), one might surmise that seeking mental health

counseling would constitute one element of active coping for a psychological disorder,

such as anorexia or depression.

However, a survey by the American Psychological Association (1996) suggested

that adults lack knowledge about how to access mental health services and what problems

are appropriately treated by a psychologist, suggesting that people may be largely

uninformed and naiYve about current psychological services and practices. In addition,

several studies found that compared to a control group, individuals labeled as seeking

mental health treatment (e.g., psychiatrist, psychiatric hospital, counseling center) were

viewed less favorably than control groups (Parish & Kappes, 1979; Phillips, 1963;

Sibicky & Dovidio, 1986). Dovidio and colleagues (1985) also found that while people

attributed some positive characteristics (e.g., competence) to individuals with problems

who sought help, they were still rated unfavorably on other dimensions (e.g., sociability)









(Dovidio et al., 1985). Seeking mental health treatment may automatically generate

negative stereotypes associated with having psychological problems or "mental illness."

For example, Piner and Kahle (1984) found that confederates who were "ex-mental

patients" were viewed as more unusual than controls, even when the confederates actual

behavior did not differ from the controls. Similarly, Phillips (1963) argued that contact

with psychiatrist or treatment at a mental hospital generates labels linked to the mentally

ill or insane, given the greater social rejection that occurred for these two help sources as

compared to a physician or clergyman. However, differences in social rejection were

much more prominent for mental disorders than for mental health treatment, suggesting

that the behaviors associated with a psychological problem may be more related to social

distancing than help-seeking itself (Phillips, 1963, 1964).


Children's Perceptions

Studies examining children's conceptions of psychological treatment have also

generated mixed findings. In a study by Dollinger and Thelen (1978), few children (10%)

had positive attitudes toward seeing a psychologist, most children's attitudes were neutral

(60%), and derogatory expressions about psychologists or people who see psychologists

were relatively rare. Notably, derogatory expressions were slightly higher among boys.

Sigleman and Mansfield (1992), found that a considerable number (68%) of youths

reported negative feelings (e.g., anxiety) toward therapy. Cross-sectional samples

involving elementary, middle, and high school youths suggest few differences as a

function of grade or age (Dollinger & Thelen, 1978; Sigelman & Mansfield, 1992).

Research involving adolescents specifically, however, suggests that some may

view psychiatric hospitalization and outpatient therapy as beneficial and understand the









risks and benefits associated with therapy, particularly if they had experience with

therapy (Kaser Boyd, Adelman, & Taylor, 1985; Pugh, Ackerman, McColgan, & de

Mesquita, 1994). While adolescents felt that psychiatric hospitalization was associated

with stigma, particularly among their peers, they considered peer stigma for outpatient

therapy a relatively rare occurrence (Pugh et al., 1994). With respect to gender, females

have been found to endorse some form of treatment for psychological problems more

often than males (Chimonides & Frank, 1998; Pugh et al., 1994).


Adolescent Help-Seeking

Research from the adolescent help-seeking literature suggests that they are often

unaware of mental health resources (Dubow et al., 1990) and prefer to seek help from

informal sources (e.g., friends and family) as opposed to more formal mental health

services (Offer & Schonert Reichl, 1992; Rickwood & Braithwaite, 1994). Further,

adolescents cite confidentiality and privacy as concerns about seeking treatment (e.g.,

(Balassone et al., 1991; Dubow et al., 1990), suggesting they have concerns about being

stigmatized should they seek psychological services.

Factors associated with adolescent help-seeking include gender, previous

exposure to therapy, ethnicity, parent education, parent marital status, and symptom level.

For example, females and those with interpersonal problems tend to hold more positive

attitudes toward seeking informal help (Boldero & Fallon, 1995; Garland & Zigler, 1994;

Schonert-Reichl, Offer, & Howard, 1995) as well as formal professional help (Rickwood

& Braithwaite, 1994). Females also seek help from friends more frequently, whereas

males tend to seek help from parents (Dubow et al., 1990).









Research suggests that African-Americans are more likely to seek out family and

friends for help rather than a mental health provider (McMiller & Weisz, 1996).

Similarly, minorities are less likely to contact a mental health professional and more

likely to wait until problems are more severe before seeking professional help (McMiller

& Weisz, 1996; Padgett, Patrick, Bums, & Schlesinger, 1994; Schonert-Reichl et al.,

1995).

Prior participation in therapy has been associated with more favorable attitudes

and increased willingness to seek help in general (Rickwood & Braithwaite, 1994).

Adolescents who have had previous psychological treatment also see therapy as more

beneficial and minimize the risks associated with it (Kaser Boyd et al., 1985; Pugh et al.,

1994).

Lower parental education, single parent status, lower self-image, lower grades,

and lower SES have also all been associated with less frequent help-seeking (Saunders et

al., 1994). Notably, many of these same variables have been associated with increased

levels of psychosocial problems among youth. These adolescents may be in double-

jeopardy-at-risk for having psychosocial problems but less likely to actually seek

services. School-based or school-linked services may be a more effective means of

reaching these youths, provided counseling with a psychologist, or other mental health

professional, is seen as a viable option.


The Role of the Peer Group

The literature further suggests that peers may play an important role in adolescent

help-seeking. Adolescents who experience psychological problems tend to seek help

from peers, rather than parents or professionals; however, those who seek informal help









are also more likely to seek formal help (Offer et al., 1991; Saunders et al., 1994). Thus,

the peer group, and particularly their prevailing attitudes toward mental health counseling

and psychological problems, could strongly encourage or discourage professional help-

seeking in adolescents. For example, adolescents with negative attitudes toward therapy

could think less of adolescents who participate in counseling, make fun of them, or

participate in fewer activities with them, thereby discouraging peers from entering

counseling. On the other hand, if adolescents hold positive views about seeking

professional help, they may encourage a peer who is having problems to seek counseling,

particularly if it is easily accessible and available at school. Adolescents could even view

the person more positively for getting the necessary help and "actively coping" with their

problem.

Given that not all mental disorders are viewed equally by adolescents (Chassin &

Coughlin, 1983; Coie & Pennington, 1976), adolescent perceptions of peers who seek

counseling may differ, depending on the nature and severity of the problem. For example,

it is doubtful that anyone would view therapy or hospitalization negatively for a person

who is suicidal or having frank delusions. However, suicidality and psychosis only

represent a small number of those who need psychological treatment. As mentioned

earlier, a considerable number of adolescents with less "severe" emotional and behavioral

disorders, such as depression or conduct disorder, do not receive the help they need. It is

for these youth that peer group attitudes toward mental disorders and therapy are

particularly important, but very little is known about whether stigma within the peer

group exists. Examining the prevailing attitudes within the peer group can provide an

important basis for intervention programming, leading to more favorable attitudes among






18


adolescents, in general, more open discussion of counseling, emotional and behavioral

problems, and, perhaps, less stigmatization and more support among the peer group.















CONCEPTUAL BASES FOR THE CURRENT STUDY


Stigma

Stigma, in the context of social psychology, represents negative stereotypes

toward a person or group of persons that can lead to prejudice or discrimination (Corrigan

& Penn, 1999). Often, the negative effect occurs as a function of the label itself, rather

than behavior. Indeed, a review of studies found that the label of "mental illness" or

"mental patient" (often with no accompanying behavioral description) results in a

negative evaluation (Hayward & Bright, 1997). But the inclusion of contextual variables,

such as actual behaviors or details of a person's life, appears to lessen negative reactions.

In this manner, attitudes toward the "label" of mental illness may not accurately reflect

attitudes toward individuals. Specifically, people's reactions to a person labeled with a

"mental illness or mental disorder" likely differ from their reactions to a person who has

an emotional and behavioral disorder but who has not been "labeled" as mentally ill. In

typical social interactions, a person observes or encounters the behavioral manifestations

of an emotional and behavioral disorder prior to having knowledge of a specific label.

Thus, most people are likely unaware of the name of the emotional and behavioral

problem a person is experiencing without someone explicitly stating it. In contrast, the

behaviors of an individual are more easily and readily observed and immediately impact

the interpersonal interaction.









The literature also suggests that, in the absence of a label, the public refrains from

labeling all but the most extreme behaviors as mental illness (Hayward & Bright, 1997).

Rather, the public tends to associate the terminology "mental illness or mental disorder"

with "severe mental illness," such as schizophrenia or other conditions that might require

psychiatric hospitalization. Much of the research literature that examines general attitudes

toward mental illness seems to reflect attitudes about severe mental illness rather than

more common emotional and behavioral disorders, such as depression and anxiety. This

study aims to go beyond measuring the stigma associated severe mental illness by using

behavioral descriptions of disorders that are commonly found among adolescents.


Attitudes

Attitudes involve the favorable or unfavorable evaluations toward stimuli

manifested in one's beliefs, feelings, or intended behavior (Myers, 1999). Essentially,

attitudes represent ways to organize information so as to react more efficiently to people,

events, and situations. In this manner, stigma might be considered a negative attitude and

measured as such. At the same time, many in social psychology have argued that attitudes

do not necessarily predict behavior in actual situations. Often this is the case when

studies restrict the measurement of attitudes to the beliefs and feelings. The lack of

relation between attitudes and behavior rings particularly true in cases where the beliefs

measured are conceptually broad but the behaviors are specific (Ajzen & Fishbein, 1977).

So, in order to increase the relationship between attitudes and actual behaviors, one

should assess beliefs about specific situations or events as well as behavioral intentions,

i.e., commitment to a future action (Siperstein, 1980). The inclusion of behavioral

intentions also suggests a degree of acceptance, or willingness to interact socially with









the specified individual. Consequently, this study employs measures that assesses

attitudes, in the form of cognitive and affective judgments, as well as acceptance, in form

of behavioral intentions.


Vignette Methodology

Vignette research methodology offers the ability to specify and standardize

information and thereby assess specific attitudes. Vignettes, which have been widely used

in attitude research, are descriptions of specific situations relevant to the constructs under

investigation (Alexander & Becker, 1978). They offer the advantage of simulating real-

life decision-making situations systematically, so as to control for extraneous factors and

randomize people to conditions (Alexander & Becker, 1978; Lanza & Carifio, 1992).

Differences in the information in vignettes serve as different social cues; evaluative

judgments and behavioral intentions in response to the vignettes are presumed to reflect

differences associated with these cues. The systematic variation of such informational

cues permits examination of effects related to one variable or combinations of variables

(e.g., one can examine effects of gender of respondent alone or in combination with

effects related to the type of emotional and behavioral problem described in the vignette).

Written vignettes are also more cost-effective and more easily standardized as compared

to videotaped presentations of actors playing different parts. Actors, themselves, might

introduce additional elements, such as general appearance, that might bias responses

related to social image. Finally, the use of vignettes may also reduce the impact of

demand characteristics, impression management, and social desirability that can bias

responses to direct questioning about attitudes (e.g., how do you feel about this person

with depression?). When asked directly about specific attitudes, people may respond in a









manner that they perceive would please the interviewer or that is socially acceptable. By

using a comparative contrastive technique, in which respondents only answer questions

about one or two potential vignettes, respondents are often unaware of the specific

information that has been manipulated (Burstin, Doughtie, & Raphaeli, 1980), potentially

reducing some social desirability bias. Although vignette methodology offers many

advantages, the stimuli presented are only simulated, and as such, generalization of

findings is limited. Actual behaviors in real situations may vary. Despite this limitation,

vignette methodology seems to provide a convenient, systematic way to assess general

attitudes among a population. The inclusion of a measure of behavioral intentions should

shed additional light on the potential link between attitudes and behaviors.















PURPOSE OF THE STUDY


The purpose of this study is to examine rural adolescents' attitudes and behavioral

intentions toward peers who have emotional and behavioral disorders and who participate

in school-based counseling with a psychologist.


Research Questions

Question 1: How do adolescents view peers with emotional or behavioral

problems?

Question 2: How do adolescents view peers who participate in counseling with a

psychologist?

Question 3: Do adolescents' views of peers change as a function of emotional

and behavioral problems displayed and counseling status?

Question 4: Do adolescents' attitudes toward peers with emotional and behavioral

problems and those who attend counseling vary as a function of gender?

Question 5: Do adolescents' attitudes toward peers with emotional and behavioral

problems and those who attend counseling vary as a function of certain other respondent

and demographic characteristics of the adolescent (e.g., perceived similarity to target

adolescents, prior counseling experience, help-seeking attitudes, grade, ethnicity, and

rurality)?








Hypotheses

The specific hypotheses outlined below are based on the existing literature and

clinical experience. The hypotheses and subsequent analysis plan are categorized as

primary or secondary. Primary hypotheses and analyses relate to the first four research

questions and concern the effects of experimental manipulation and the gender of the

respondent/target peer. The secondary hypotheses and analyses relate to the last research

question and concern the potential moderating effects of respondent and demographic

characteristics on adolescent ratings.


Primary Hypotheses: Experimental Manipulation

1. Adolescents will view peers who display emotional or behavioral problems more

negatively than peers with no reported problems. Previous research suggests that

younger children view emotionally-disturbed peers less favorably than medically-ill,

handicapped, or learning disabled youth (Novak, 1974; Parish et al., 1978; Roberts et

al., 1984). The same pattern likely exists within an adolescent sample.

2. Adolescents will view peers with conduct disorder more negatively than peers with

depression or peers reporting high family conflict. Compared to those with other

disorders, children with externalizing difficulties and behavior problems tend to be

viewed less favorably (Coie & Pennington, 1976; Hoffman et al., 1977; Marsden et

al., 1977). Furthermore, youth with conduct disorder are likely to exhibit behaviors

that are more salient and problematic than those who experience depression or family

conflict. Indeed, aggression, one component of conduct disorder, is one of the

strongest predictors of rejection by the peer group (Coie et al., 1990). Adolescents

may also attribute more personal responsibility and a lack of self-control to









adolescents exhibiting conduct problems. As such, conduct disorder likely elicits less

sympathy than either depression or family conflict and would be viewed more

negatively.

3. Among peers who have no apparent emotional and behavioral problems, adolescents

will view peers who participate in counseling less favorably than those who do not

participate in counseling. No known studies have examined children's or

adolescents' attitudes toward peers who participate in counseling independent of the

reason for counseling. Research with adults, however, suggests more negative views

of individuals who seek mental health treatment as compared to individuals without

apparent problems who do not attend therapy (Dovidio et al., 1985; Parish & Kappes,

1979; Piner & Kahle, 1984; Sibicky & Dovidio, 1986). Further, stigma may be

associated with mental health treatment by proxy. In the absence of information about

the type of problem, an adolescent may assume that a person attending counseling

must be experiencing some type of mental health problem and subsequently attribute

a negative stereotype, such as those that have been associated with "severe mental

illness."

4. Among peers who have emotional and behavioral problems, adolescents will view

peers who attend counseling more favorably than peers who do not attend

counseling. Although behavioral manifestations and labels of emotional and

behavioral disorders have negative connotations, individuals who seek counseling for

these problems have been viewed as more competent (Dovidio et al., 1985). In

addition, individuals seen as actively coping with their problem (i.e., doing something

to alleviate the problem) may be viewed more favorably (Schwarzer & Weiner,









1991). As such, it is expected that adolescents will view peers who participate in

counseling as actively coping with their problems and competent, and as a result,

view peers with emotional and behavioral problems more favorably if they are

participating in counseling.

5. Adolescents' perceptions of peers who participate in counseling will also differ as a

function of the type of emotional and behavioral problem displayed. Given evidence

that adolescents can discriminate among disorders (Chassin & Coughlin, 1983; Coie

& Pennington, 1976; Offer & Schonert-Reichl, 1991), their perceptions of counseling

may also vary depending on the disorder. For example, depression may be viewed as

more internal to the person, biological, or as the result of situational factors.

Compared to conduct disorder, depression may elicit more sympathy than conduct

disorder and be viewed as a more valid reason for attending counseling. Alternatively,

adolescents may perceive family conflict as a normative experience rather than a

disorder, per se and thus, may view these peers less favorably for attending

counseling for such a normative problem (Sigelman & Mansfield, 1992). Finally,

although conduct disorder may be viewed as highly problematic and in need of

treatment, peers with conduct disorder may be viewed so negatively that participation

in counseling may not mitigate these views.

6. Among peers who participate in counseling, adolescents will view peers with

apparent emotional and behavioral problems more favorably than peers without

apparent emotional and behavioral problems. Knowing the reason for counseling

(i.e., that the person is experiencing a specific emotional or behavioral problem) may

mitigate negative attitudes toward peers who participate in counseling. In the absence









of explicit problems, adolescents might presume that an adolescent who attends

counseling must have "serious problems" and attach the general label of "mental

illness" or "mental disorder" it the peer, which would have negative connotations

(Corrigan & Penn, 1999; Hayward & Bright, 1997). Consequently, it expected that

the peers with emotional and behavioral problems who attend counseling will be

viewed as actively coping with their problems and be viewed more favorably than

peers who do not appear to have a valid reason (i.e., no apparent emotional and

behavioral problems) for attending counseling.

7. Males will view same-sex peers with depression more negatively, whereas females

will view same-sex peers with conduct disorder more negatively. Emotional problems,

such as depression, are more common and seem to be more acceptable among

females; whereas behavior problems, such as conduct disorder, are more common

among males (Ostrov et al., 1989) and include behaviors that are less acceptable for

females. Consequently, we expect that males will view peers with depression more

negatively. Likewise, we expect females may view peers who exhibit overt conduct

problems less favorably. Family conflict, however, is more pervasive across genders,

and thus, no differences are expected between male and female ratings of peers with

family conflict.

8. Females will view same-sex peers who participate in counseling more positively

compared to males. Previous research has found that females hold more favorable

attitudes toward help-seeking (Boldero & Fallon, 1995; Garland & Zigler, 1994),

acknowledge their own need for help (Saunders et al., 1994), and endorse treatment









for a wider range of problems than males (Pugh et al., 1994). Consequently, it is

likely that females will view peers who seek help more favorably than males.

9. Males will view same-sex peers who are depressed and attend counseling most

negatively; whereas, females will view same-sex peers who are depressed and attend

counseling most positively. Given that males are likely to view other males who are

depressed more negatively and given their less favorable attitudes toward help-

seeking (Boldero & Fallon, 1995; Garland & Zigler, 1994; Schonert-Reichl et al.,

1995), it is expected that males would rate their same-sex peers who are depressed

and attend counseling least favorably. For females, the opposite may be true. Females

may view other females with depression more positively than other disorders, and

given females' generally positive attitude toward help-seeking, it is expected that

females with rate their same-sex peers with depression who attend counseling most

positively.


Secondary Hypotheses: Respondent and Demographic Characteristics

Respondent characteristics

10. Adolescents who have had prior experience with counseling will view peers who

participate in counseling more positively than adolescents who have had no prior

experience with counseling. In adolescence, personal participation in mental health

treatment has been associated with increased willingness to seek help, stronger

endorsement of help-seeking for emotional and behavioral problems, increased

perceptions of benefits and decreased perceptions of risks associated with mental

health treatment (Kaser Boyd et al., 1985; Pugh et al., 1994; Rickwood &









Braithwaite, 1994). Consequently, adolescents with prior counseling experience are

expected to provide more favorable ratings of peers who participate in counseling.

11. Adolescents who have more positive attitudes toward help-seeking will view peers

who participate in counseling more favorably than adolescents with more negative

views of help-seeking. If adolescents who hold positive views about help-seeking for

themselves, it is expected that they will also view peers who attend counseling more

favorably than adolescents who have lower help-seeking attitudes.

12. Adolescents who perceive themselves to be more similar to the target peer in the

vignette will view that peer more positively than adolescents who do not perceive

themselves as similar to the target peer described in the vignette. Perceived similarity

suggests that a rater perceived the other person as having common interests or traits,

which would result in more favorable ratings. In addition, perceived similarity can

also increases a person's empathy and lead one to rate the person more favorably.


Demographic Characteristics

13. Compared to minority youth, Caucasian youth will have the most positive attitudes

toward peers who attend counseling. Research indicates that ethnic minorities, and

African-American youth in particular, tend to seek formal help less frequently, wait

until problems are more severe before seeking help, and rely more on informal

sources of help (McMiller & Weisz, 1996; Padgett et al., 1994; Schonert-Reichl et al.,

1995). Consequently, ethnic minority youth are expected to view peers who attend

counseling less favorably.

14. Older adolescents will view peers with emotional and behavioral problems or who

participate in counseling more favorably than younger adolescents. Older children









have a better understanding of emotional and behavioral disorders given their

increasing cognitive maturity, entry into formal operations stage, and decreased

egocentrism (Coie & Pennington, 1976; Dollinger, Thelen, & Walsh, 1980; Sigelman

& Mansfield, 1992). Likewise, they are better able to appreciate others' perspectives

and understand causes of disorders. As a consequence, it is expected that older

adolescents will hold more favorable views of peers with emotional and behavioral

problems. The literature also suggests that adolescents are also able to understand

better the risks and benefits of counseling better than children (Kaser Boyd et al.,

1985). Consequently, older adolescents may view peers who participate in counseling

more favorably than younger adolescents.

15. Compared to adolescents who live "in town," adolescents who live "in the country"

will have less favorable views toward peers who participate in counseling. The

stigma associated with mental health treatment appears to be more prevalent in rural

areas (Berry & Davis, 1978; Kenkel, 1986). Conceptions of mental health problems

and treatment tend to be associated with more severe forms of mental illness, such as

schizophrenia, and residents of rural areas tend to be less familiar with therapy as a

means of coping with less severe emotional and behavioral problems. Consequently,

it is expected that adolescents living in more rural areas would view peers who

participate in therapy less favorably than youth living in town.















METHOD


Experimental Conditions

The independent variables manipulated in this study included type of emotional

and behavioral problem (depression, conduct disorder, family conflict, no problems) and

counseling status (attending counseling vs. not attending counseling). Given that the

previous literature suggests gender differences related to help-seeking and prevalence

rates for certain disorders, gender was also as included an independent variable. Vignettes

(see Appendix A) were modeled after vignettes previously used with elementary and

middle school youth (e.g., Morgan & Wisely, 1996). All vignettes began with

information introducing the peer as a new student in school and providing other non-

specific information. Michael and Erica were chosen as the names of the peers in the

vignettes since both names can be found among thie primary ethnic populations of the

school (i.e., Caucasian and African-American). All vignettes, except the ones in which

the target peer did not have an emotional and behavioral problem, then described the peer

as missing school a lot recently and having his or her grades go down, which could be

potential outcomes for any of the emotional and behavioral problems. The outcome

information was standardized in an attempt to approximate severity across problem

conditions. Finally, specific information about emotional and behavioral problems and

counseling status was included in the vignette (except for the vignette without emotional

and behavioral problems).









Information presented in the vignettes described one of three types of emotional

and behavioral problems: conduct disorder, depression, and family conflict, or no

problems. The three types of problems selected are among the most common emotional

and behavioral problems experienced by adolescents (Dubow et al., 1990; Kashani et al.,

1987; Whitaker et al., 1990) and are among those encountered frequently in the school's

mental health service program (Danda, Evans, Rey, & Nitzberg, 2000). Reviews of

DSM-IV criteria, clinical case studies, and clinical experience provided the basis for the

behavioral descriptions of each of the emotional and behavioral problems. Descriptions

for each problem type contained approximately four behaviors consistent with common

and easily observed symptoms of that problem. Vignettes with descriptions of emotional

and behavioral problems were then presented to two psychologists and four advanced

graduate students in order to ensure validity of the symptom descriptions. All raters

independently and successfully identified each of the emotional and behavioral problem

conditions described in the vignette (i.e., depression, conduct disorder, family conflict),

indicating that the behavioral descriptions used in the vignettes accurately depicted the

behavioral manifestations of each of the emotional and behavioral problem types targeted

in this study.

Information in the vignettes also varied according to whether the target adolescent

attended counseling with a psychologist who worked at their school. The term

"counseling" was employed rather than therapy to be more consistent with common

terminology, as indicated by school personnel. At the same time, the wording "with a

psychologist" was included in order to differentiate the nature of mental health

counseling from other forms of counseling (e.g., guidance counseling). Finally, the









phrase "who works at your school" was included to reflect the growing numbers of

psychologists offering mental health services within the school setting, including the

school district in which this study took place.


Measures

Dependent Measures

Attitudes

The Adjective Checklist (Siperstein, 1980; Siperstein & Gottlieb, 1977) assesses

children's judgments of the attributes toward an actual or hypothetical peer (see

Appendix B). The measure includes 32 adjectives, half of which are negative and half of

which are positive. Subjects circle as many or as few of the adjectives they feel best

describe the target child. The original Adjective Checklist included the adjective

"handsome", which was changed to "good-looking" in order to use terminology more

applicable to both genders.

Scores are obtained by subtracting the negative adjectives from the positive

adjectives and adding a constant of 20 in order to eliminate negative scores (Friedrich,

Morgan, & Devine, 1996; Siperstein, 1980). Siperstein (1980) suggested that scores

below 20 represent an overall negative evaluation of the target child since more negative

than positive adjectives were chosen as descriptors, while scores over 20 represent an

overall positive evaluation of the target child since more positive than negative adjectives

were chosen as descriptors. The measure has been used extensively to assess school-age

children's attitudes about peers with mental retardation and physical handicaps (e.g., Bak

& Siperstein, 1986; Gottlieb & Gottlieb, 1977; Morgan & Wisely, 1996; Siperstein &

Gottlieb, 1977; Wisely & Morgan, 1981). Factor analysis has confirmed the construct









validity of the positive and negative valences of the adjectives, and the measure has

demonstrated good internal consistency reliability (Cronbach's alpha = .81; Siperstein,

1980).


Behavioral intentions

The Adolescent Activity Preference List was based on the Activity Preference

Scale (Siperstein, 1980; see Appendix C), which was designed to assess children's

behavioral intentions toward an actual or hypothetical peer. The participant rates how

willing they are to engage in a variety of common social activities with a target child. The

original Activity Preference List contained 30 items that were derived from a list of

activities generated by middle and upper elementary school youth. A shorter, 15-item

version demonstrated excellent reliability (Cronbach's alpha = .90; Siperstein, 1980).

Similar activity preference measures have been used with elementary and middle school

children to measure attitudes towards physically or mentally handicapped children (e.g.,

(Bak & Siperstein, 1986; Morgan & Wisely, 1996; Wisely & Morgan, 1981).

A new Adolescent Activity Preference List (see Appendix D) was developed to

include more developmentally and geographically appropriate items. The original 30

items from the Activity Preference Scale plus 13 new items were pilot tested with three

adolescents (two males, one female) who rated the appropriateness of the activity,

indicated the frequency of occurrence for the list of activities, and provided comments

about the activities listed. Items that were consistently rated as occurring frequently (e.g.,

daily or once a week by at least 2 of the 3 pilot respondents) and deemed

developmentally appropriate were retained for the new scale. The final version of the

Adolescent Activity Preference List included 19 items. Responses are rated on a 4-point









scale, anchored by the following phrases: like a lot, like, dislike, and dislike a lot. Higher

scores indicate less willingness to participate in activities with the target peer.


Manipulation check

Five true-false questions were presented after the vignette to assess accurate

encoding of information about independent variables manipulated and to provide a

cursory indication of reading comprehension (see Appendix E). The first two questions

pertained to basic information present in all vignettes. The next three questions pertained

to behavioral indicators of the emotional and behavioral problem and counseling status,

and varied according to the vignette. In order to be retained in the study analyses,

participants had to correctly answer at least 2 of the 3 questions related to the

experimental manipulation and correctly answer at least 3 of the 5 total questions.


Respondent Characteristics

Help seeking

The Help Seeking Scale (Garland & Zigler, 1994) is a 26-item measure designed

to assess children's and adolescent's attitudes willingness to seek help for psychosocial

problems from adults in the school setting (See Appendix F). The measure was modeled

after the Attitudes Towards Seeking Professional Help Scale (Fischer & Turner, 1970)

and Propensity to Seek Help Scale (Kessler, Reuter, & Greenley, 1979, as cited in

Garland & Zigler, 1994), which were both designed to measure adult attitudes. In order to

assess attitudes toward seeking professional help more accurately, the word "teacher"

was replaced with "therapist". Scores range from 26 to 94, with higher scores indicating

more positive attitudes towards seeking help for emotional and behavioral problems. The

measure has demonstrated good internal consistency (.80- .85) and test-retest reliability









(.81 .89) (Garland & Zigler, 1994). Items asking about actual and prior help-seeking are

not scored in the total.


Prior experience with therapy/counseling

One question, "Have you ever talked to a therapist or counselor about a personal

or emotional problem?" regarding the individual's prior experience with help-seeking

was already included as part of the Help Seeking Scale. In order to provide a broader

definition of previous exposure to counseling/therapy, two additional questions were

added: Has a family member or relative ever talked to a therapist or counselor about a

personal or emotional problem?; Has one of your friends ever talked to a therapist or

counselor about a personal or emotional problem?"


Social desirability

The Help Seeking Scale included five yes/no items from the Minnesota

Multiphasic Personality Inventory (MMPI) that assess social desirability. Given the

possibility of social desirability influence in attitude research, these items were retained

for use in analyses. A composite score was generated by coding "1" for socially-desirable

responses and "0" for non-socially desirable responses and then by summing the items to

form a scale. Higher scores indicate greater social desirability.


Perceived similarity

Adolescents were asked to rate how much the person in the vignette was like him

or her. Responses were based on a 4-point scale, anchored by the phrases not at all like

me, not much like me, somewhat like me, and very much like me.









Demographics

Participants were asked to list their age, grade, ethnicity, and gender at the

beginning of the Help-Seeking Scale. In addition, participants were asked to indicate

whether they lived "in town" or "in the country" in order to provide a measure of rurality.


Procedure

Given the age of the participants, anonymous nature of the study, and survey

procedure, the Institutional Review Board granted the study exempt status (i.e.,

completion of questionnaires indicated consent to participate). After subsequent approval

from the local school board and principal, 20 classrooms were randomly chosen to

participate in the study. Math and English classes were excluded from randomization at

the request of school personnel. Randomization of classrooms was completed in blocks

according to grade level to increase the likelihood of equal distribution of experimental

conditions across grade level. However, many classrooms available for participation in

the study contained students in multiple grade levels, and no classrooms contained

exclusively 12th graders. No information was available at the time about the gender

distribution of the classrooms. The number of classrooms selected was ultimately

determined by the number of students needed (approximately 500) to ensure adequate

power for the proposed analyses.

During a faculty meeting, the principal investigator (C.E.D.) described the study

and notified teachers whose classrooms were selected for data collection. All teachers

agreed to participate; however, two classrooms were unavailable on the day of data

collection. One class could not participate due to a previously planned activity and agreed

to have data collected the following week. The other class was misidentified as meeting









during the class period designated for data collection, so no further data could be

gathered.

Two weeks before the anticipated day of data collection, a letter to the student's

parents (see Appendix G) that explained the study was sent home with each student who

attended a class designated for data collection. Only one parent requested that her child

not participate, and this student worked quietly at his/her desk while other students

completed the study questionnaires.

Trained research assistants administered questionnaire packets and provided

verbal instructions to all students in participating classrooms. Questionnaire packets also

included a cover sheet (see Appendix H) explaining the study and emphasizing voluntary

participation along with one vignette and the remaining questionnaires. Questionnaire

completion was voluntary and anonymous (i.e., no identifying information was

collected), and students who did not wish to participate were allowed to work quietly at

their desks. After students turned in their completed packets, students received a

debriefing statement (see Appendix I) that explained the purpose of the study in more

detail and provided contact information should they have questions or concerns.

Each student received a questionnaire packet with one of 8 vignette descriptions

about a hypothetical new student who was the same gender as the student. The gender of

the target peer in the vignette was matched to the gender of the respondent, and packets

were color-coded for ease of administration (i.e., males received blue packets that

contained a vignette with a male target peer, and females received green packets that

contained a vignette with a female target peer). Each of the 8 conditions (Emotional and

Behavioral Problem x Counseling Status) was sequentially distributed to students within








each gender to ensure equal sample sizes across conditions. Thus, each student only

answered questions about one vignette in which the peer was the same gender as the

student. The first two questionnaires, which related to the vignette specifically, were

counterbalanced to avoid order effects.


Participants

Participants were 453 adolescents attending grades 9-12 at Columbia High School

(CHS), which is located in a rural county of north Florida. Of the 453 packets returned,

six were deleted from analyses due to missing or invalid data, and five were deleted due

to incorrect responses on the manipulation check, leaving a total of 442 participants in the

study. No formal data was collected regarding blank packets returned (indicating refusal

of the adolescent to participate), but it appeared that only a few adolescents per class did

not participate.

Table 1 lists the demographic characteristics of the sample. Participants included

significantly more females than males (! (1) = 7.61, p < .001). The gender distribution

(48% males and 52% females) for 10-12th graders for the CHS 1999-2000 school year

suggests that females may be slightly over-represented in this sample. Significantly more

Caucasian than minority students participated (J (1) = 65.06, p < .001); however, the

racial distribution within the sample is roughly equivalent to the current racial

distribution within the high schools (70% Caucasian and 25% African-American).

Significantly fewer 12th graders participated in the study relative to other grade levels (3!

(3,439) = 48.17, p < .001). The small percentage of 12th graders in the study likely

reflects the lack of classes containing exclusively 12th graders. Slight differences related

to rurality also emerged LX! (1,441) = 3.27, p = .07)], with a little over one-third of









students characterizing themselves as living in town, while almost half of the students

characterized themselves as living in the country. However, approximately one-fifth of

students did not answer this question. Due to concerns about the students' ability to

report about their socioeconomic status, participants did not respond to any questions

about their family's income or professional background. Data from the Columbia County

School District reported that 52% of students are eligible for free or reduced lunches,

suggesting that at least half the students in this school district are in a lower

socioeconomic bracket.

Table 2 presents the sample size across the experimental conditions. Chi-square

analyses revealed that despite gender differences in overall response rate, the proportion

of males and females was equally distributed among the counseling status [W (1) = 1.45,

P = .23)] and type of emotional and behavioral problem [_ (3) = .411, p = .94)]

conditions. Further chi-square analyses revealed that each of the 8 conditions across

gender contained relatively equal proportions of students LX (15) = 12.59, 12 = .63)], with

no significant differences across ethnic [X2 (15) = 10.94, p = .76)], grade [& (45) = 32.92,

p = .91)], or rural [& (15) = 18.97, p = .21)] distributions.












Table 1. Demographic Characteristics of Participants (N=442).
Frequency Percentage
Gender
Male 192 43.4%
Female 250 57.6%


Ethnicity
Caucasian
African-American
Other
Missing Data

Rurality
Living in town
Living in the country
Missing Data


Grade


9
10
11
12
Missing Data


Table 2. Subject Distribution across Experimental Conditions.

Counseling No Counseling
Emotional and
Behavioral Problem Male Female Male Female Total

No Problem 23 33 27 29 112

Depression 29 32 20 33 114

Conduct Disorder 25 28 19 35 107

Family Conflict 27 28 22 32 109

Total 104 121 88 129 442


304
96
49
3


160
194
88


120
108
155
54
5


68.8%
21.7%
8.9%
0.6%


36.2%
43.9%
19.9%


27.1%
24.4%
35.1%
12.2%
1.2%















ANALYSIS PLAN


Questionnaires were visually inspected to detect random responding, and

participants with random responses or who failed the manipulation check were deleted

from all analyses. Psychometric properties were examined for each of the dependent

measures. For this study, an individual's scores on the Adolescent Activity Preference

List and the Help Seeking Scale were calculated by summing across all items the

individual completed and dividing this total by the number of items that individual

completed. In this way, calculation of the total score takes into account missing data as

well as yields a more interpretable total score (i.e., scores correspond to valences

attributed to the anchor phrases for each of the 4 points). In order to determine whether

students answered in a socially-desirable manner, correlations between social desirability

and the dependent measures were generated; if social desirability correlated with either

dependent measure, it was to be analyzed as a covariate in subsequent analyses.

Correlations respondent and demographic characteristics were also examined.

A 4 (Emotional and Behavioral Problem Type) x 2 (Counseling Status) x 2

(Gender of Target Peer/Respondent) ANOVA was conducted to determine main and

interaction effects of the experimental conditions and gender. Analyses were conducted

separately for each of the two dependent measures: the Adjective Checklist and the

Adolescent Activity Preference List: Adjective checklists tap into a primarily cognitive

component (internalized ideas and evaluation), whereas measures of behavioral intentions









tap into a primarily connotative component (behavioral commitment) (Gottlieb &

Gottlieb, 1977). Several studies using similar measures have also found stronger effects

for the adjective checklist as well as small to insignificant correlations between adjective

checklists and measures of behavioral intentions (Siperstein & Gottlieb, 1977; Wisely &

Morgan, 1981). In light of this previous research, it was hypothesized that the Adjective

Checklist would produce more significant and stronger effects than the Adolescent

Activity Preference List.

Significant univariate main effects and interactions were followed up by multiple

comparisons. Given the number of multiple comparisons, Sidak's test was employed to

control for experimentwise error. The Sidak test [1 (1 e) 1] is similar to the

Bonferroni correction, in that the alpha is adjusted for the number of tests performed, but

tends to be less stringent.

Secondary analyses included variables related to respondent characteristics (i.e.,

Prior Counseling Experience with counseling/therapy, Help Seeking attitudes, and

Similarity to target adolescent) and demographic variables (i.e., Ethnicity, Grade level,

and Rurality). In order to determine the impact of these variables as related to the

experimental conditions, each of the six variables was added separately and

independently to the primary ANOVA model (i.e., Emotional and Behavioral Problems,

Counseling Status, and Gender). Again, a separate ANOVA was conducted for each of

the dependent variables with significant univariate main effects and interactions followed

up by multiple comparisons, using Sidak's test to control for experimentwise error.

In order to maximize sample sizes across the experimental conditions, several

levels of the variables used in the secondary analyses were combined. For Prior









Counseling Experience, two variables were created-one involving a composite yes/no

variable based on an endorsement of any one of the three questions related to prior

experience (self, family, or friend) and the other involving the single question asking if

the adolescent had talked to a therapist or counselor about a personal or emotional

problem. Analyses were run with the composite Prior Experience variable first, with the

plan that if no significant effects emerged, the single question Prior Experience variable

would be utilized. For the purposes of the ANOVA analyses, help seeking attitudes were

categorized as "Favorable" or "Unfavorable" based on their total score. Since values up

to 3.00 corresponded with the negative valence of "disagreeing" with a statement, all

total scores up to 3.00 were considered "Unfavorable" attitudes. Likewise, scores greater

than 3.00 were considered "Favorable" since they corresponded the positive valence of

"agreeing" with a statement. As for the similarity to target adolescent in the vignette,

scores of 1 or 2 (not at all like me, not much like me) were grouped as "Not Similar" and

scores of 3 and 4 (somewhat like me, very much like me) were grouped as "Similar."

Given the small sample size of other minorities relative to African-American and

Caucasian groups, only the latter two groups were included in analyses related to

ethnicity.















RESULTS


Data Inspection

Questionnaires that had greater than 20% of items missing were considered

invalid, but participants were retained if they had completed at least one of the dependent

measures (the Adjective Checklist or the Adolescent Activity Preference List)

sufficiently. Six packets that had suspect or invalid data were removed (e.g., circled

female but completed male version of questionnaire, did not complete either dependent

measures sufficiently, or created own categories and made inappropriate comments). Five

additional questionnaires were removed due to a failure to answer enough manipulation

items correctly. As a result of deletion of invalid or suspect data, 442 (97.6%) of the

original 453 participants were available for subsequent analyses.

Of the 442 participants, nine participants (2.4%) either did not complete the

Adolescent Activity Preference List or were deleted from subsequent analyses for that

measure due to excessive missing data. Of the remaining 433 valid Adolescent Activity

Preference List questionnaires, thirty (6.9%) had 3 or fewer items missing and were

included in subsequent analyses. On the Adjective Checklist, fourteen participants (3.1%)

did not complete the questionnaire, leaving a total of 428 valid questionnaires for use in

analyses involving that measure. On the Help Seeking Scale, eleven participants (2.5%)

did not complete the measure or were deleted from subsequent analyses due to excessive

missing data. Of the remaining 431 valid Help Seeking Scale questionnaires, 34 (7.9%)









had 3 or fewer items missing and were included in subsequent analyses. The scoring of

the Adolescent Activity Preference List and the Help Seeking Scale allowed for

imputation of missing data through summing the total for each participant and dividing

the total score by the number of items answered.

Examination of the frequencies and distributions of the dependent measures

identified five outliers (i.e., scores beyond 3 standard deviations from the mean) on the

Adjective Checklist and one outlier on the Help Seeking Scale. These scores fell along

the continuum of scores in the distribution (i.e., there were no significant gaps between

the outliers and the rest of the scores). Further, visual inspection of these participants'

response patterns on other questionnaires revealed no other anomalies. As such, these

individuals' scores were not considered invalid outliers and were retained in subsequent

analyses.


Psychometric Analyses

Histograms of the frequency of participants' scores on the dependent measures as

well as the Help Seeking Scale suggested that the distributions of these measures

followed a normal curve. In addition, measures of skewness (i.e., symmetry of

distribution) and kurtosis (i.e., peakedness or flatness of distribution) were all within the

1 range, further supporting the presence of a normal distribution across these measures.

Item-total correlations and alpha reliability coefficients were calculated for each

dependent measure and the Help Seeking Scale. Internal consistency of the dependent

measures ranged from adequate on the Adjective Checklist (a = .75) to excellent on the

Adolescent Activity Preference List (a = .96), and were consistent with findings from

previous literature (Siperstein, 1980). The Help Seeking Scale also demonstrated good









internal consistency (a = .85), consistent with previous literature (Garland & Zigler,

1994). On the other hand, the 5 items comprising the social desirability scale

demonstrated poor internal consistency (_ = .35), suggesting that items did not reliably

assess social desirability. However, the mean score for social desirability items was to .71

(S = .91), with a median and mode of 0, indicating that participants generally did not

answer in a socially desirable manner. Since few respondents endorsed any of the items

in a socially-desirable manner, it is likely that there was random scatter across items

endorsed in a socially-desirable manner may account for the poor reliability of the scale.


Descriptive Statistics

Scores on the Adjective Checklist ranged from 4.0 to 36.0 (higher scores indicate

more positive attitudes toward the target peer), with an average of 19.41 (SD = 5.75).

Scores above 20 indicate overall favorable attitudes, and scores below 20 indicate overall

unfavorable attitudes. Scores on the Adolescent Activity Preference List ranged from 1.0

to 4.0 (higher scores indicate lower acceptance), with an average score 2.31 (SD = .58).

Tables 3 and 4 present the means and standard deviations for the Adjective Checklist and

the Adolescent Activity Preference List, respectively, for each of the 8 conditions across

gender. The Adjective Checklist and the Adolescent Activity Preference List were

somewhat correlated (r = -.36, p < .001), indicating that as perceptions of the target peer

improved, willingness to engage in activities with the target peer also improved slightly.

Correlations between social desirability and the dependent measures were not significant

(Adjective Checklist, r = .007, p = .89; Adolescent Activity Preference List, r = .03, p =

.50), so social desirability was not included as a covariate in subsequent analyses.








Pearson correlations between the six respondent and demographic variables revealed only

one significant correlation (Similarity and Help Seeking; r = -.15; R < .002).

Table 5 presents respondent characteristics of the participants. With respect to

Perceived Similarity, only 28.5% of the adolescents viewed themselves as similar to the

target adolescent described in the vignettes. Significantly fewer adolescents (20% vs.

54%) viewed themselves as similar to the target peers described with an emotional and

behavioral problems as compared to the target peer without any problems [& .(1) = 46.24,

p < .001)]. Sixty-one percent of the sample reported having sought help from a counselor

or therapist themselves and/or knowing someone who had sought such help, with

approximately one-quarter of the sample reported having sought help in the past

themselves. There were no gender differences related to prior counseling experience or

perceived similarity to the target peer.

Scores on the Help-Seeking Scale ranged from 1.39 to 3.83, with an average of

2.76 (SD = .40). Thus, adolescents' views of help seeking in this sample were slightly

unfavorable but approached favorable ratings (i.e., score greater than 3.0). Females had

significantly higher help-seeking scores (M_= 2.83) than males (M = 2.66) [F (1,429) =

19.52; p< .001]. When scores on the Help Seeking Scale were categorized as Favorable

vs. Unfavorable (for the purposes of the ANOVA analysis), 69% of scores fell within the

unfavorable attitudes category. Again, a significantly higher proportion of females (75%

vs. 25%) had favorable attitudes relative to males [& (1,430) = 21.72, p < .001)]. When

asked to endorse any individual, to whom they would turn if sad or upset, adolescents

overwhelmingly endorsed friends (86%) and parents (52%). Two-fifths of the sample

would turn to a sibling, and 28% would turn to another adult. Only 10% and 6% would








seek out a teacher or counselor, respectively, if sad or upset. Finally, 15% indicated that

they might not seek help from any one.


Primary ANOVA Analyses

Results of the ANOVA for the Adjective Checklist revealed that the 4 (Emotional

and Behavioral Problems) x 2 (Counseling Status) x 2 (Gender of Respondent/Target

Peer) model reached significance [F (15,412) = 25.21, 1 < .001, E2 = .48]. Likewise, the

same ANOVA model for the Adolescent Activity Preference List reached significance [F

(15,417) = 8.49, p < .001, e2 = .23]. Main effects and interactions are described below.


Main Effects

Emotional and Behavioral Problem

A significant main effect for emotional and behavioral problems, which

corresponded to Hypotheses 1 and 2, emerged on both the Adjective Checklist IF (3,412)

= 102.05, p <.001, e2 = .43] and Adolescent Activity Preference List [F (3,417) = 19.28,

S< .001, 2 = .12]. Multiple comparison analyses revealed full support for Hypothesis 1

(i.e., adolescents would view peers with emotional and behavioral problems more

negatively than peers without such problems) on the Adjective Checklist, but only partial

support for Hypothesis 1 on the Adolescent Activity Preference List. Means for each

dependent measure are presented in Table 6.

On the Adjective Checklist, as seen in Figure 1, adolescents rated target peers in

the No Problem condition (M = 25.26) significantly more positively than target peers in

the three emotional and behavioral problem conditions: Depression (M = 17.63; p < .001;

ES = .60), Conduct Disorder (_M = 15.52; p < .001; ES = .73), and Family Conflict (M =








18.67; 12 < .001; ES = .59). Further, target peers described as having emotional or

behavioral problems obtained average scores less than 20, indicating that more negative

than positive adjectives were chosen as descriptors (i.e., an overall unfavorable attitude).

In contrast, target peers described as not exhibiting symptoms of emotional and

behavioral problems obtained average scores greater than 20, indicating that more

positive than negative adjectives were chosen as descriptors (i.e., an overall favorable

attitude). On the Adolescent Activity Preference List, slightly different findings emerged.

As seen in Figure 2, adolescents' responses indicated a significantly higher willingness to

participate in activities with peers in the No Problem condition (M = 2.29) compared to

the Conduct Disorder condition (M = 2.67; p < .001; ES = .33). The No Problem

condition did not differ from the Depression or Family Conflict conditions.

Multiple comparison analyses further revealed full support for Hypothesis 2 (i.e.,

peers in the Conduct Disorder condition would be viewed more negatively than peers in

the other two emotional and behavioral problem conditions) for both dependent

measures. As predicted, target peers in the Conduct Disorder were rated significantly

more negatively than target peers in the Depression (p < .002; ES = .25) or Family

Conflict (p < .001; ES = .40) condition on the Adjective Checklist. Likewise, adolescents

were significantly more willing to engage in activities with target peers in the Depression

(M = 2.20; p < .001; ES = .38) and Family Conflict (M = 2.20; p < .001; ES = .39)

conditions relative to peers in the Conduct Disorder condition (ML = 2.67).

In sum, a main effect for Emotional and Behavioral Problems versus No Problems

was shown only for the Adjective Checklist (Hypothesis 1); however, adolescents








consistently rated peers with Conduct Disorder the least favorably and were least willing

to participate in activities with them (Hypothesis 2).


Counseling Status

Analyses revealed no significant main effects for Counseling Status [F (1,412) =

.047, = .828] on the Adjective Checklist or the Adolescent Activity Preference List [F

(1,417) = .068, p = .794].


Gender of Respondent/Target Peer

Although not predicted, a significant main effect emerged for Gender on the

Adjective Checklist [F (1,412) = 22.64, p < .001, e2 = .05] and the Adolescent Activity

Preference List [F (1,417) = 55.82, p < .001, E 2= .12]. On the Adjective Checklist,

females (M = 20.26) rated same-sex peers in all vignettes significantly more positively

than males rated their same-sex peers (M = 18.28; p < .001; ES = .14). Similarly, females

(M = 2.14) were significantly more willing than males (M = 2.53; p < .001; ES = 32) to

participate in the activities enumerated on the Adolescent Activity Preference List with

their same-sex peers. Thus, females in this study generally tended to describe their same-

sex peers more positively than males and were more willing to engage in activities with

these peers than males.


Interactions

Emotional and Behavioral Problem x Counseling Status

A significant Emotional and Behavioral Problem x Counseling Status interaction

effect, which corresponded to Hypotheses 3 through 6, emerged on the Adjective

Checklist [F (3,412) = 4.812, p < .003, 2 = .03], but not on the Adolescent Activity









Preference List [F (3,417) = .053, p = .984]. Estimated marginal means on the Adjective

Checklist for the Emotional and Behavioral Problem x Counseling Status interaction are

listed in Table 7 and depicted in Figure 3. Results of multiple comparison analyses

revealed full support for Hypothesis 3 (i.e., for peers with no apparent emotional and

behavioral problems, those who participated in counseling would be viewed less

favorably as compared to those who did not participate in counseling). Specifically, target

peers in the No Problem, Counseling condition (M = 23.99) received significantly lower

ratings than target peers in the No Problem, No Counseling condition (M = 26.53 <

.002). Notably, ratings of target peers described with an emotional and behavioral

problem remained, on average, below 20 (i.e., an overall unfavorable attitude), whereas,

target peers described without a emotional and behavioral problems averaged higher than

20 (i.e., an overall favorable attitude), regardless of counseling status. Further,

examination of effect sizes associated with decreases in attitudes suggested that the

effects of counseling status (ES = .21) was not as strong as the overall effect of having an

emotional and behavioral problem (ES = .64).

Hypothesis 4 (i.e., among peers with emotional and behavioral problems, those

who were in counseling would be viewed more favorably than those who were not in

counseling) and Hypothesis 5 (i.e., differences between ratings of peers in counseling

would differ based on the type of emotional and behavioral problem) received little

support. Although differences between ratings of target peers with Conduct Disorder in

the No Counseling condition (M = 14.79) and Counseling condition (M = 16.26) were in

the expected direction and approached significance (p = .087; ES = .17). At the same

time, the small increase in attitudes toward peers with conduct disorder who attended









counseling did not greatly improve their ratings relative to peers in the Depression (M =

18.18; V = .11; ES = .22). and Family Conflict.(M = 18.83; p < .01; ES = .35) conditions.

Thus, there was little overall change in attitudes toward peers with emotional and

behavioral problems associated with counseling (Hypothesis 5).

Hypothesis 6 (i.e. for adolescents participating in counseling, those who had

obvious emotional and behavioral problems would be viewed more favorably than peers

who did not have obvious emotional and behavioral problems) did not receive support.

Attending counseling neither greatly improved that status of target peers with emotional

and behavioral problems, nor greatly lessened the status of target peers without apparent

symptoms of emotional and behavioral problems.


Gender x Emotional and Behavioral Problem

A significant Emotional and Behavioral Problems x Gender interaction effect,

which corresponded to Hypothesis 7, emerged on the Adjective Checklist [F (3,412) =

3.43, p < .02, e2 = .02], but not on the Adolescent Activity Preference List LF (1,412) =

.126, R = .723]. Multiple comparison analyses revealed full support for Hypothesis 7 for

females (i.e., females would view same-sex peers with Conduct Disorder most

negatively) but only partial support for males (i.e., males would view same-sex peers

with Depression most negatively). Estimated marginal means on the Adjective Checklist

for the Emotional and Behavioral Problems x Gender interaction are listed in Table 8 and

depicted in Figure 4.

As expected, females rated same-sex peers in the Conduct Disorder condition (M

= 15.72) significantly lower than same-sex peers in either the Depression (M = 19.04; p <

.001; ES = .37) or Family Conflict (M = 19.17; p < .001; ES = .47) conditions. In partial








support of Hypothesis 7, males rated same-sex peers in the Depression condition (M =

16.21) significantly lower than same-sex peers in the Family Conflict (M = 18.18; p <

.001; ES = .25), but not the Conduct Disorder (M = 15.32) condition. Unexpectedly, male

targets in the Conduct Disorder condition were rated significantly lower than male targets

in the Family Conflict condition (p < .001; ES = .32) but did not differ significantly from

male targets in the Depression condition. The interaction described above was likely

facilitated by the fact that females rated their same-sex peers with Depression (M =

19.03) significantly higher relative to males (M = 16.21, p < .001; ES = .30). Contrary to

expectations, no such gender effect emerged for the peers in the Conduct Disorder

condition.


Gender x Counseling Status

No significant Gender x Counseling Status interaction effect, which corresponded

to Hypothesis 8, emerged on either the Adjective Checklist [F (1,412) = .126, p = .723]

or the Adolescent Activity Preference List [F (1,417) = .027, P = .868]. Thus, no

evidence was found suggesting that females view same-sex peers in counseling more

positively than males.


Emotional and Behavioral Problem x Counseling Status x Gender.

A significant Emotional and Behavioral Problem x Counseling Status x Gender

interaction effect, which corresponded to Hypothesis 9, emerged on the Adjective

Checklist [F (3,412) = 3.124, p < .03, 2= .02] but not on the Adolescent Activity

Preference List [F (3,417) = 1.97, p = .119]. Means and Standard Deviations for all

conditions are presented in Table 1. The means are graphed depicted in Figure 5.

Multiple comparisons revealed partial support for Hypothesis 9 (i.e., males would view









same-sex peers with depression who attend counseling most negatively, whereas females

would view same-sex peers with depression who attend counseling most positively).

Specifically, males tended to view same-sex peers in the Conduct Disorder and

Family Conflict conditions slightly more positively in the Counseling condition relative

to the No Counseling condition (p = .06; ES = .31 and .19; ES = .20, respectively). No

significant effect was observed for male target peers in the Conduct Disorder condition.

As a result, males did not rate their same-sex peers in the Depression, Counseling

condition the lowest of all conditions as predicted. Rather, males rated same-sex peers in

the Counseling condition significantly lower in the Depression condition as compared to

the Family Conflict condition (p < .04; ES = .40), but relatively equivalent to those in the

Conduct Disorder condition (ES = .05), thereby providing only partial support for the

first part of Hypothesis 9.

As expected, females rated same-sex peers in the Depression condition,

significantly higher in the Counseling condition as compared to the No Counseling

condition (P < .005; ES = .30). Likewise, within the Counseling condition, female targets

in the Depression condition were rated significantly more positively than in the Conduct

Disorder condition (p < .001; ES = .46), and moderately more positively in the Family

Conflict condition (p = .10; ES = .21), supporting the second half of Hypothesis 9.

Notably, female target peers in the Depression, Counseling condition obtained scores

slightly higher than 20, indicating an overall favorable attitudes from female

respondents-one of the few times that peers with an emotional and behavioral problem

were viewed favorably overall.









In sum, Hypothesis 9 was fully supported relative to females but only partially

supported relative to males. Males and females had significantly stronger and opposite

reactions to their same-sex peers with depression, depending on counseling status.

Specifically, females viewed their same-sex peers with depression in counseling most

positively, and males viewed their same-sex peers with depression less positively relative

to those with family conflict but not conduct disorder.


Secondary ANOVA Analyses

Prior Experience with Counseling/Counseling

Hypothesis 10 postulated that adolescents who had prior experience with

counseling would view peers who participated in counseling more favorably (i.e.,

Counseling Status x Prior Experience interaction). ANOVA analyses revealed no

significant effects on either dependent measure for Prior Experience defined as a

composite of having sought help in the past themselves and/or knowing someone who

has sought help from a counselor or therapist. When Prior Experience was redefined to

include only individuals who had themselves sought help from a therapist or counselor,

several significant main effects and interactions emerged on the Adolescent Activity

Preference List, but not the Adjective Checklist. Yet, neither the proposed Counseling x

Prior Experience interaction that corresponded directly to Hypothesis 10 nor a three-way

interaction that included Counseling and Prior Experience was significant. Thus,

Hypothesis 10 was not supported. The significant main effects and interactions that were

not predicted but did emerge are described below.

A significant main effect for Prior Experience emerged on the Adolescent

Activity Preference List [F (1,392) = 21.58, p <.001, E2 = .05]. Adolescents who








reported having seen a therapist or counselor in the past were significantly more willing

to engage in activities with the target peer (M = 2.13) than adolescents who had no

history of prior counseling or counseling experience (M = 2.41, 2 < .001).

ANOVA revealed a moderately significant Emotional and Behavioral Problem x

Gender x Prior Experience interaction [F (3, 392) = 2.58, p = .053, 2 = .02] on the

Adolescent Activity Preference List. Estimated marginal means for the Adolescent

Activity Preference List are presented in Table 9 and depicted in Figure 6. Significant

gains related to Prior Experience were found for females rating same-sex peers in the

Conduct Disorder (p < .001) and Family Conflict (p < .01) conditions, but not the

Depression and No Problem conditions. On the other hand, significant or nearly

significant gains related to Prior Experience were found for males rating same-sex targets

in the No Problem (p = .07), Depression (p < .001), and Family Conflict conditions (P =

.054), but interestingly enough, not the Conduct Disorder condition.

The largest increases between adolescents with No Prior Experience and those

with Prior Experience were found for males rating male targets with depression and

females rating female targets with conduct disorder. Ultimately, these differences resulted

in changes in males' and females' willingness to participate in activities among the

disorders. Specifically, males with No Prior Experience remained significantly less

willing to engage in activities with male targets in the Conduct Disorder condition

relative to those in the Family Conflict condition (p < .01). The significantly large

increase in status for male targets with Depression who were rated by males with Prior

Experience (and lack of change for those with Conduct Disorder) resulted in males' being

significantly less willing to engage in activities with male targets in the Conduct Disorder









as compared to the Depression condition (p < .01). Similarly, females with No Prior

Experience were significantly less willing to engage in activities with female targets in

the Conduct Disorder condition than those in all three other conditions: No Problem (R <

.001), Depression (p < .001), or Family Conflict (p < .001) conditions. However, these

differences disappeared for females with Prior Experience, largely accounted for by the

significant increase in status for female targets with Conduct Disorder who were rated by

females who had Prior Counseling experience.


Help-Seeking Attitudes

Analyses revealed no significant Counseling x Help Seeking Attitudes interaction,

which corresponded to Hypothesis 11 (i.e., adolescents who had more favorable attitudes

toward help seeking would view adolescent targets in counseling more positively than

adolescents with unfavorable attitudes toward help seeking). However, a significant main

effect emerged on the Adolescent Activity Preference List [F (1,390) = 10.45, p < .002,

E2 = .03]. Adolescents who had more favorable attitudes toward help seeking were

significantly more willing to engage in activities with target peers (M = 2.19) than

adolescents with unfavorable attitudes (M = 2.41; p < .001), regardless of counseling

status and emotional and behavioral problems.


Similarity

A main effect for Similarity emerged on the Adjective Checklist [F (1,396) =

4.38, p < .04, E2 = .01] and the Adolescent Activity Preference List [F (1,401) = 13.31, p

< .001, E2 = .03], which provided full support for Hypothesis 12 (i.e., adolescents who

perceived themselves as similar to the target peer would rate the target peer more








favorably than adolescents who did not perceive themselves as similar to the target

adolescent). Adolescents who perceived the target peer in the vignettes as Similar to

themselves rated the target peer significantly higher (M = 19.45) on the Adjective

Checklist than adolescents who perceived the target peer as Not Similar to them (M =

18.85; p < .04). Likewise on the Adolescent Activity Preference List, adolescents who

perceived the target peer in the vignettes as Similar were significantly more willing to

share activities with the target peer (M = 2.18) than adolescents who perceived the target

peer as Not Similar (M = 2.41; p < .001).

In addition to the main effect, several significant three-way interactions emerged.

Specifically, a significant Emotional and Behavioral Problems x Counseling Status x

Similarity to Target interaction emerged [F (3, 396) = 2.94, p < .03, E2 = .02] on the

Adjective Checklist. In addition, a three-way interaction between Emotional and

Behavioral Problems, Gender, and Similarity to Target approached significance [F (3,

396) = 2.33, p = .074, E2 = .02] on the Adjective Checklist, and achieved significance on

the Adolescent Activity Preference List [F (3, 396) = 2.73, p < .05, e2 = .02].

Estimated marginal means for the Emotional and Behavioral Problems x

Counseling Status x Similarity to Target on the Adjective Checklist are presented in

Table 10 and graphed in Figure 7. Primarily, the effects of perceived Similarity and

Counseling Status differed in the Depression condition. Specifically, if adolescents

perceived the target peer in the Depression condition as Similar, they rated target peers in

the No Counseling condition significantly higher (p < .05), but rated target peers in the

Counseling condition significantly lower (p < .01). Likewise, adolescents who perceived

the target peer as Not Similar rated target peers in the Depression condition significantly









higher in the Counseling condition relative to the No Counseling condition (p < .004). In

contrast, adolescents who perceived the target peer as Similar to them rated target peers

in the Depression condition significantly lower in the Counseling relative to the No

Counseling condition (p < .04).

Estimated marginal means for the Emotional and Behavioral Problems x Gender x

Similarity to Target interaction on the Adolescent Activity Preference List are presented

in Table 11 and graphed in Figure 8. The primary interaction effects related to Similarity

were evident in the Female, Conduct Disorder conditions and the Male, Family Conflict

conditions. Females who perceived the target female as Similar were significantly more

willing to engage in activities with same-sex peers in the Conduct Disorder condition (P

< .004). Notably, the significantly lower ratings for target females in Conduct Disorder

relative to the other three conditions (all p < .003) when target female was perceived as

Not Similar disappeared if the target female was perceived as Similar. On the other hand,

males who perceived themselves as Similar to the target male were moderately

significantly more willing to engage in activities with target males in the Family Conflict

condition (p <008) and No Problem (p = .07) conditions.


Ethnicity

Hypothesis 14 proposed that Caucasian adolescents would view target peers in

counseling more positively than ethnic minority adolescents (i.e., Ethnicity x Counseling

interaction). Limitations in sample size restricted the analyses to comparisons between

African-Americans and Caucasians. No significant interactions related to ethnicity and

counseling emerged on either dependent measure. Thus, Hypothesis 14 was not

supported.









Analyses did reveal a significant main effect for Ethnicity on the Adjective

Checklist [F (1,386) = 3.75, p = .054, E2 = .01], such that African-American youth gave

slightly higher ratings (M = 20.07) than Caucasian youth (M = 19.03) overall. In

addition, a significant Ethnicity x Emotional and Behavioral Problems interaction [F (1,

393) = 2.90, p = .09, e2 = .01] emerged on the Adolescent Activity Preference List.

Estimated marginal means are presented in Table 12 and graphed in Figure 9. African-

American youth (M = 2.05) were more willing to engage in activities with peers in the

No Problem condition relative to Caucasian youth (M = 2.34; p < .02). In addition,

multiple comparisons between the emotional and behavioral problem conditions among

each ethnic group revealed slightly different patterns. For African-American youth, the

pattern was similar to the pattern originally found (i.e., they were less willing to

participate in activities with peers in Conduct Disorder condition relative to the other

three conditions). Caucasian youth, on the other hand, were significantly more willing to

participate in activities with peers in the Depression condition (M = 2.12; p < .04) but

significantly less willing to participate in activities with peers in the Conduct Disorder

condition (p <002), relative to the No Problem condition. (M = 2.34). Again, peers in the

Conduct Disorder condition were seen less favorably than those in the other two

emotional and behavioral problem conditions (p < .001).


Grade Level

Hypothesis 15, which corresponded to Emotional and Behavioral Problems x

Grade and Counseling Status x Grade interactions, postulated that older adolescents

would view peers with emotional and behavioral problems and those who participated in

counseling more favorably than younger adolescents. No significant main effects or








interactions emerged on the Adolescent Activity Preference List, and no significant main

effects or two-way interactions emerged on the Adjective Checklist. However, ANOVA

analyses revealed a significant Counseling Status x Gender x Grade interaction [F (3,

359) = 2.75, V < .05, E2 = .02] on the Adjective Checklist. Estimated marginal means are

presented in Table 13. Contrary to expectations, males in 12th grade rated same-sex peers

in the Counseling condition significantly lower than peers in the No Counseling condition

(V < .02). Thus, Hypothesis 15 was not confirmed. However, given the small sample size

of 12th graders in this study, this finding may be spurious and would need further

replication before generalizing to other populations.


Rurality

Hypothesis 16, which corresponded to a Counseling Status x Rurality interaction,

postulated that adolescents in more rural areas would have more negative views toward

peers who participate in counseling. A Counseling Status x Rurality interaction emerged

on the Adjective Checklist [F (1, 392) = 4.80, p < .03, E2 = .02], providing support for

Hypothesis 16. Estimated marginal means are presented in Table 14 and graphed in

Figure 10. Within the Counseling condition, adolescents who lived in Town (M = 20.29)

had a moderately significant tendency to rate peers more favorably as compared to

adolescents who lived in the Country (M = 19.13; p = .08). Somewhat surprisingly,

adolescents who lived in Town tended to rate peers in the Counseling condition (M =

20.29) higher than peers in the No Counseling condition (M = 18.77; p = .07).

Interestingly, their rating was also over 20, indicating that adolescents living in town held

overall favorable attitudes of adolescents who participated in counseling. In contrast, no

differences emerged related to rurality and emotional and behavioral problems.











Table 3. Means and Standard Deviations across Experimental Conditions on the
Adjective Checklist (N=434)
No Therapy Therapy
Emotional and
Behavioral Problem Male Female Male Female

No Problem
M 24.63 28.43 22.18 25.80
SD 4.47 4.14 7.37 3.78
Depression
M 16.60 17.55 15.83 20.53
SD 4.60 4.29 3.90 5.27
Conduct Disorder
M 14.06 15.51 16.59 15.93
SD 3.62 3.15 4.24 3.37
Family Conflict
M 17.37 19.67 19.00 18.67
SD 4.65 3.61 3.23 2.99

Note. Scores range from 4 (most negative) to 36 (most positive). Scores over 20 indicate
overall favorable attitudes, and scores below 20 indicate overall negative attitudes.


Table 4. Means and Standard Deviations across Experimental Conditions on the
Adolescent Activity Preference List (N = 433)
No Counseling Counseling
Emotional and
Behavioral Problem Male Female Male Female

No Problem
M 2.42 2.14 2.63 1.95
SD .52 .41 .62 .38
Depression
M 2.39 1.99 2.40 2.00
SD .46 .56 .62 .55
Conduct Disorder
M 2.83 2.46 2.83 2.56
SD .58 .52 .51 .57
Family Conflict
M 2.44 1.96 2.30 2.08
SD .53 .43 .59 .44

Note. Scores range from 1 (most positive) to 4 (most negative).











Table 5. Respondent Characteristics of Participants Across Gender (N = 442).
Frequency Total Percentage
of Sample
Pearson Chi
Male Female Sar
Square

Similarity to Target Peer .48
Similar 58 68 28.5%
Not Similar 134 182 71.5%

Prior Experience with
Counseling/Therapy .26
Self 49 61 24.9%
2.28
Self, Family, and/or Friend 112 169 61.1%

Attitude Toward Help-Seeking' 21.72*
Favorable 31 93 28.0%
Unfavorable 152 155 69.5%

tBased on Help Seeking Score total; N = 431
*P <.001


Table 6. Means on the Adjective Checklist and Adolescent Activity Preference List for
Emotional and Behavioral Problems Main Effect


Emotional and Behavioral
Problem


Adiective Checklist


Adolescent Activity
Preference List


No Problem 25.26 2.29

Depression 17.63 2.20

Conduct Disorder 15.52 2.67

Family Conflict 18.67 2.20

Note. Scores on the Adjective Checklist range from 4 (most negative) to 36 (most
positive). Scores over 20 indicate overall favorable attitudes, and scores below 20
indicate overall negative attitudes. Scores on the Adolescent Activity Preference List
range from 1 (most positive) to 4 (most negative).





























Figure 1. Main Effect of Emotional and Behavioral Problems on the Adjective
Checklist


Figure 2. Main Effect for Emotional and Behavioral Problems on the Adolescent
Activity Preference List


35

30

25

n20

15

10

5

0


Conduct Disorder Family Conflict


No Disorder Depression


Conduct Disorder Family Conflict


No Disorder


1

1.5

2

2.5

3

3.5

4


Depression


I | I 0











Table 7. Estimated Marginal Means on the Adjective Checklist for Emotional and
Behavioral Problems x Counseling Status Interaction
Emotional and Behavioral
Problem No Counseling Counseling
No Problem 25.53 23.99

Depression 17.01 18.18

Conduct Disorder 14.79 16.26

Family Conflict 18.51 18.83

Note. Scores range from 4 (most negative) to 36 (most positive). Scores over 20 indicate
overall favorable attitudes, and scores below 20 indicate overall negative attitudes


4-No Problem ,-Depression
-*-Conduct Disorder -*)K-Family Conflict


28-


I


22

120]

18

16

14

12


.1


No Counseling


Counseling


Figure 3. Interaction Effect for Emotional and Behavioral Problems x
Counseling Status on the Adjective Checklist


I











Table 8. Estimated Marginal Means on the Adjective Checklist for Emotional and
Behavioral Problems x Gender Interaction
Emotional and Behavioral Male Female
Problem
No Problem 23.41 27.11

Depression 16.21 19.04

Conduct Disorder 15.32 15.72

Family Conflict 18.18 19.17

Note. Scores range from 4 (most negative) to 36 (most positive). Scores over 20 indicate
overall favorable attitudes, and scores below 20 indicate overall negative attitudes


-- -No Problem -1- -Depression
-A-Conduct Disorder -*-Family Conflict

28

26 -

24

22

20

18 -

16

14-

12

10 -,
Male Female


Figure 4. Interaction Effect for Emotional and Behavioral Problems x
Gender on the Adjective Checklist










-4--No Disorder -*- Depression
-A- Conduct Disorder -)K-Family Conflict


30
28
26
24
22


18
16
14
12
10


No
Counseling Counseling
Male Male


No
Couseling
Female


Counseling
Female


Figure 5. Interaction Effect for Emotional and Behavioral Problems x Counseling
Status x Gender on the Adjective Checklist (see Table 3 for Means)


I I I I I











Table 9. Estimated Marginal Means on the Adolescent Activity Preference List for
Emotional and Behavioral Problems x Gender x Prior Counseling Experience Interaction.
Male Female
Emotional & No Prior Prior No Prior Prior
Behavioral Problem Experience Experience Experience Experience
No Problem 2.63 2.33 2.05 2.02

Depression 2.56 2.05 2.00 2.00

Conduct Disorder 2.84 2.71 2.66 2.12

Family Conflict 2.46 2.10 2.11 1.72

Note. Scores range from 1 (most positive) to 4 (most negative).


---No Problems -- Depression
-4- Conduct Disorder --- Family Conflict

No Prior Prior No Prior Prior
Experience Experience Experience Experience
Male Male Female Female
1.5- '---'-


1.75-2z

2- 0


2.25 -

2.5- -

2.75-

3 -


Figure 6. Interaction Effect for Emotional and Behavioral Problems x Prior
Counseling Experience x Gender on the Adolescent Activity Preference List












Table 10. Estimated Marginal Means on the Adjective Checklist for Emotional and
Behavioral Problems x Counseling Status x Perceived Similarity Interaction.
No Counseling Counseling
Emotional &
Behavioral Problem Not Similar Similar Not Similar Similar
No Problem 25.19 27.28 22.86 25.86

Depression 16.43 19.48 19.00 15.88

Conduct Disorder 14.67 14.25 16.18 16.60

Family Conflict 17.89 20.42 18.39 19.83

Note. Scores range from 4 (most negative) to 36 (most positive). Scores over 20 indicate
overall favorable attitudes, and scores below 20 indicate overall negative attitudes

-4-No Disorder -U- Depression
S-i-Conduct Disorder -)-Family Conflict


No No
Counseling Counseling
Not Similar Similar


Counseling
Not Similar


Counseling
Similar


Figure 7. Interaction Effect for Perceived Similarity x Emotional and Behavioral
Problems x Gender on the Adjective Checklist








Table 11. Estimated Marginal Means on the Adolescent Activity Preference List for
Emotional and Behavioral Problems x Gender x Perceived Similarity Interaction
Male Female
Emotional &
Behavioral Problem Not Similar Similar Not Similar Similar
No Problem 2.68 2.41 2.16 1.98

Depression 2.45 2.24 2.05 1.86

Conduct Disorder 2.84 2.82 2.61 2.00

Family Conflict 2.49 2.05 2.02 2.10

Note. Scores range from 1 (most positive) to 4 (most negative).



-4- No Problem 4- -Depression
--Conduct Disorder --<-Family Conflict


Not Similar Similar
Male Male


1.75


2


2.25


2.5


2.75


3


Not Similar
Female


Figure 8. Interaction Effect for Perceived Similarity x Emotional and Behavioral
Problems x Gender on the Adolescent Activity Preference List


Similar
Female






72




Table 12. Estimated Marginal Means on the Adolescent Activity Preference List for
Ethnicity x Emotional and Behavioral Problems Interaction
Emotional and
Behavioral Problem Caucasian African-American
No Problem 2.34 2.05

Depression 2.12 2.31

Conduct Disorder 2.63 2.84

Family Conflict 2.16 2.23

Note. Scores range from 1 (most positive) to 4 (most negative).


-- No Problem -- Depression
-A- Conduct Disorder -- Family Conflict


Caucasian


1


1.5


2


2.5


3


3.5

4


African-American


Figure 9. Interaction Effect for Ethnicity x Emotional and Behavioral Problem on
the Adolescent Activity Preference List











Table 13. Estimated Marginal Means on the Adjective Checklist for Counseling Status x
Gender x Grade Interaction
No Counseling Counseling

Grade Male Female Male Female

9 17.31 20.09 18.57 20.52

10 18.59 20.92 19.22 19.13

11 17.52 20.43 18.73 20.45

12 21.25 19.31 16.18 20.66

Note. Scores range from 4 (most negative) to 36 (most positive). Scores over 20 indicate
overall favorable attitudes, and scores below 20 indicate overall negative attitudes











Table 14. Estimated Marginal Means on the Adjective Checklist for Rurality x
Counseling Status Interaction.
Counseling
Status In Town In the Country

No Counseling 18.67 19.90

Counseling 20.29 19.14

Note. Scores range from 4 (most negative) to 36 (most positive). Scores over 20 indicate
overall favorable attitudes, and scores below 20 indicate overall negative attitudes


-i-No Counseling -U-Counseling

24

23

22

21

20-

19-

18

17-

16-
In Town In the Country

Figure 10. Interaction Effect for Rurality x Counseling Status on the Adjective
Checklist















DISCUSSION


The primary goal of this study was to examine adolescents' attitudes toward

common emotional and behavioral disorders and school-based counseling with a

psychologist in a rural community. Five specific questions were addressed: (1) How do

adolescents view peers with emotional or behavioral problems? (2) How do adolescents

view peers who participate in counseling with a psychologist? (3) Do adolescents' views

of peers change as a function of emotional and behavioral problems displayed and

counseling status? (4) Do adolescents' attitudes toward peers with emotional and

behavioral problems and those who attend counseling vary as a function of gender? (5)

Do adolescents' attitudes toward peers with emotional and behavioral problems and those

who attend counseling vary as a function of certain other respondent and demographic

characteristics of the adolescent (e.g., perceived similarity to target adolescents, prior

experience with counseling, help-seeking attitudes, grade, ethnicity, and rurality)?

Attitudes were measured via an adjective checklist and a measure of behavioral

intentions.


How Do Adolescents View Peers with Emotional and Behavioral Problems?

Adolescents had consistently less favorable attitudes toward peers described as

exhibiting behaviors associated with depression, conduct disorder, and family conflict as

compared to peers without such problems. On average, adolescents endorsed more

negative and fewer positive descriptors for target peers described as having emotional









and behavioral problems; whereas adolescents endorsed more positive and fewer negative

descriptors for target peers without such problems. However, on the measure of

behavioral intentions, the negative effect of having an emotional or behavioral problem

was only evident for peers with conduct disorder. Adolescents indicated less acceptance,

as measured by willingness to engage in social activities, of peers who displayed

conduct-disordered behaviors as compared to peers without apparent problems. In

contrast, no differences in willingness to participate in activities were observed between

peers without problems and those with depression or family conflict.

Since the descriptions of depression, family conflict, and conduct disorder

included somewhat negative behaviors (e.g., not wanting to hang out with peers as much,

arguing, fighting, getting upset more easily), these are likely to invoke negative

perceptions, which could be easily indicated by some of the negative descriptors included

on the Adjective Checklist. However, on the measure of behavioral intentions,

adolescents did not evidence exclusionary behavior toward peers with depression and

family conflict. Granted, individuals experiencing family conflict and depression may not

be as "fun" for some activities, but based on the results of this study, they do not appear

to be prone to exclusion from activities. Thus, despite somewhat negative attitudes

toward peers with emotional and behavioral problems, these attitudes seem to have less

of an effect on adolescents' willingness to interact with these peers. Peers with conduct

disorder were the exception.

Adolescents consistently rated conduct-disordered peers significantly less

favorably and were less willing to engage in social activities with them as compared to

peers with family conflict, depression, or no problems. These findings are consistent with









previous research with young children suggesting that children who exhibit externalizing

problems, and aggression in particular, are less well-liked (Coie & Pennington, 1976;

Hoffman et al., 1977; Marsden et al., 1977) and frequently rejected (Coie et al., 1990).

Indeed, the behaviors described within the conduct disorder vignettes (e.g., arguing,

fighting, losing temper) are commonly associated with peer rejection in adolescence as

well (Inderbitzen-Pisaruk & Foster, 1990).

Adolescents' increased cognitive ability also allows them to consider differences

among emotional and behavioral problems relative to causal attributions, locus of control,

and impact on others, which, in turn, could result in differential reactions to peers

exhibiting symptoms of emotional and behavioral problems. For example, adolescents

may attribute more personal responsibility (i.e., a person chooses to behave in such a

manner or at least has more control over such behavior) for conduct-disordered behaviors

as compared to behaviors associated with depression or family conflict. Likewise,

problems such as depression and family conflict might be viewed as more external to the

person, and as such, related to situational events and circumstances rather than personal

choices. Such externally-based problems can inherently elicit more sympathy,

understanding, and tolerance, and have less negative influence on adolescents'

willingness to socialize with these peers. In addition, behaviors commonly exhibited by

peers with conduct disorder (e.g., fighting, skipping school, stealing, etc.) often violate

social norms and could result in negative consequences for both the individual and his or

her peers. Specifically, adolescents, who are prosocial as a group and concerned about

conformity to the peer group (regardless of gender), may be less willing to associate with

peers who could get them into trouble. It is important to remember, however, that despite









their overall negative evaluation by the peer group, youth with conduct disorder are not

necessarily friendless. In fact, rejection of aggressive and deviant peers by the prosocial

or mainstream group frequently leads them to associate with similarly deviant peers and

continued antisocial behavior. A group of antisocial youth may be rejected by the

mainstream crowd, but the deviant peer group itself can provide acceptance and

validation for conduct-disordered youth and their behaviors.


How Do Adolescents View Peers Who Attend Counseling with a Psychologist?

The impact of counseling on adolescents' attitudes was evident across the No

Problems condition. Adolescents attributed less favorable attitudes toward peers without

apparent problems if they participate in counseling. Thus, in the absence of any other

negatively-valenced information (such as problem behaviors), the mere mention of

counseling with a psychologist generated more negative opinions of these peers. In this

respect, seeking mental health treatment may automatically generate negative stereotypes

associated with having psychological problems, which has been suggested by previous

research (Phillips, 1963; Piner & Kahle, 1984). Perhaps, peers attribute negative

stereotypes, such as "only weak or dependent people need help" or "a person must have

be really messed up to need counseling," when they do not know the reason for

counseling.

At the same time, peers who attended counseling for no apparent reason still

obtained significantly more favorable ratings than peers who exhibited behaviors

associated with emotional and behavioral problems, regardless of whether or not they

were receiving counseling. This finding suggests that the absolute magnitude of stigma

related to the counseling versus no counseling conditions appears less than the degree of









stigma generated by the presence of emotional and behavioral problems, per se. In this

respect, these results provide support for the primacy of stigma for emotional and

behavioral problems relative to the stigma for counseling itself.

Alternatively, the relative lack of findings related to counseling may also reflect

the location of psychological treatment and the treatment terminology used in the study.

For example, more negative attitudes have also been found for individuals who have been

characterized as seeing a psychiatrist or as having been hospitalized for mental illness

(Nunally, 1961; Phillips, 1963). The term "counseling" may have less negative

connotations than the term "therapy." Certainly, clinical experience with this population

and the staff at the high school seem to suggest that the term "counseling" is more widely

used and accepted and that the term "therapy" can elicit more negative connotations.

Further, counseling with a psychologist, particularly one embedded within the school

system, may trigger less negative stereotypes given the familiarity of location,

particularly as compared to stereotypes that might be associated with being treated or

medicated by a psychiatrist or with being hospitalized in a psychiatric ward.


Do Adolescents' Views of Peers Change as a Function of Emotional and Behavioral
Problems Displayed and Counseling Status?

Contrary to expectations, perceptions of peers with emotional and behavioral

problems did not improve greatly as a function of counseling status. Although

adolescents had a slight tendency to rate target peers with conduct disorder more

favorably if they attended counseling, the improved ratings did not significantly increase

their status relative to peers with other emotional and behavioral problem conditions.

Thus, peers with conduct disorder were still viewed most negatively independent of

counseling status. Even though participation in counseling did not substantially improve









ratings of target peers with emotional and behavioral problems, there was no evidence to

support greater negative attitudes as a function of attending counseling. This finding

further supports that idea that stigma, in the form of negative evaluation, associated with

emotional and behavioral problems is stronger than the stigma for attending counseling.

In other words, adolescents may already have somewhat negative attitudes toward peers

based on the behaviors they are exhibiting and knowledge of peers' participation in

counseling seems to have little detrimental effect, and potentially some positive effect, on

these negative attitudes. Although this finding may appear to be in contrast with previous

research suggesting that individuals who seek help may be viewed more positively, or at

least more competently (Dovidio et al., 1985), the measures used in this study assessed

more global attitudes. Had this study used a multi-dimensional measure of attitudes and

impressions, different results may have ensued.


Do Adolescents' Attitudes Toward Peers with Emotional and Behavioral Problems or
Who Attend Counseling with a Psychologist Vary as a Function of Gender?

Gender produced significant effects, primarily in attitudes toward emotional and

behavioral problems. It should be noted that the gender of the target peer matched the

gender of the respondent; females always rated female target peers, and males always

rated male target peers. Thus, effects related to females' perceptions of males and vice

versa could not be examined. In general, female participants responded more positively to

target peers across all vignette conditions relative to male participants, consistent with

previous research (Lopez, 1991). Contrary to expectations, females did not view their

same-sex peers with conduct disorder less favorably than males. At the same time, female

peers in the conduct disorder condition were rated the least favorably compared to the

other emotional and behavioral problem conditions. This finding is expected given the









lower prevalence rates and stereotypically non-normative nature of externalizing

behaviors for females. For similar reasons, i.e., the lower prevalence rates and the

stereotypically non-normative nature of internalizing problems for males, it was expected

that males would rate their same sex peers with depression most negatively. It was also

expected that males would have more positive ratings of peers with conduct disorder than

females, given its higher prevalence rate and the more stereotypically normative nature of

externalizing behaviors for males. Males rated their same-sex peers with depression

lower relative to females' ratings of females with depression. However, males did not

view their same-sex peers with conduct disorder any more favorably than females viewed

females with conduct disorder. Males viewed same-sex peers with both conduct disorder

and depression lower relative to peers with family conflict. It may be that the

externalizing behaviors described (stealing and skipping school) in the vignettes are more

severe than aggressive behaviors (such as fighting and asserting oneself) that might be

more accepted among males. These findings again underscore the pervasive dislike of

peers who have substantial externalizing problems.

While no gender differences emerged related for counseling per se, males and

females did view counseling somewhat differently depending upon the emotional and

behavioral problem condition. As compared to peers in the no counseling conditions,

male participants gave more favorable ratings to same-sex target peers in family conflict

and conduct disorder conditions who participated in counseling. It appears that males

who experience depression and attend counseling could have "two strikes" against them

when it comes to male peer perceptions. Already, males seem somewhat less accepting of

peers with depression. Since they generally tend to have less favorable views of help-









seeking as well, it may be even more unacceptable if peers receive counseling for a

problem that is not well-accepted. In contrast, males also may be more tolerant and

understanding toward externalizing behaviors that could be somewhat more normative

and stereotypical, and family conflict, which also tends to be more pervasive among

adolescents. As such, males may be more understanding or cognizant of the potential

value of counseling for these problems and deem these problems as more acceptable for

counseling.

As expected, females participants gave higher ratings to same-sex target peers

with depression who participated in counseling. In fact, females rated same-sex target

peers with depressive symptoms who attended counseling the most favorably among the

three emotional and behavioral problem conditions, and their ratings indicated a more

favorable attitude overall. Female relationships tend to be more intimate and disclosing in

nature than male relationships, and females tend to have more favorable attitudes toward

help-seeking (Boldero & Fallon, 1995; Garland & Zigler, 1994; Schonert-Reichl et al.,

1995). Consequently, they may have more positive attitudes towards discussing

problems with someone else, particularly for peers with depression, which appears to be a

more accepted problem among females.

Thus far, the results of this study provide support for stigma attached to behaviors

commonly associated with depression, conduct disorder, and family conflict. Differences

in adolescents' attitudes as a function of gender and counseling were less prominent.

Despite these differences, however, attitudes toward peers with emotional and behavioral

problems consistently remained below that of peers with no apparent problems, with the

exception of depressed females who attended counseling. Although attending counseling









without an apparent reason resulted in slightly unfavorably attitudes, the attitudes of

peers without emotional and behavioral problems remained significantly above that of

peers with emotional and behavioral problems, regardless of counseling status. Overall,

there was little evidence that counseling negatively impacted attitudes for peers who have

emotional and behavioral problems, and in some instances, attending counseling actually

improved attitudes slightly, particularly for females with depression.

Study results were more evident on the measure assessing evaluative judgments

of peers as compared to the measure of behavioral intentions. Conduct disorder was the

exception, which had significant effects on both dependent measures. The general lack of

correspondence between the dependent measures likely reflects the small correlation

between attitudes (i.e., cognitive and affective evaluation) and behaviors; attitudes do not

always translate into behaviors (Ajzen & Fishbein, 1977). In this case, poor

correspondence may be a good result. Specifically, the results suggest that negative

evaluations may not significantly impact willingness to interact with youths who have

emotional and behavioral problems, at least in relation to general social activities. It

should be noted, however, that many of the activities listed in the measure of behavioral

intentions were relatively impersonal and not reserved for close friendships. Findings

from previous studies suggest that youth are less willing to befriend people with mental

disorders or engage in more personal, intimate activities (Lopez, 1991; Reetz &

Shemberg, 1985). In this respect, it remains unclear whether adolescents would initiate

friendships with youth exhibiting symptoms of emotional and behavioral problem, or

how adolescents might behave or react toward close friends who have emotional and

behavioral problems.










Do Adolescent Demographic and Respondent Characteristics Affect Their Attitudes
Toward Peers with Emotional and Behavioral Problems or Who Attend Counseling with
a Psychologist?

Additional analyses examining the impact of respondent and demographic

characteristics revealed another possible explanation for the lack of correspondence

between the measures. Specifically, willingness to interact with peers differed as a

function of the personal characteristics of the respondent in combination with counseling

status or the behaviors associated with emotional and behavioral problems. In other

words, characteristics related to having an emotional and behavioral problem or attending

counseling may result in a negative evaluation for a peer, but it is characteristics of the

adolescents themselves that seem to determine their actual willingness to interact socially

with the peer. For the most part, the personal characteristics of the adolescents moderated

effects related to emotional and behavioral problems rather than counseling status,

contrary to original expectations.

For example, females who had prior counseling experience or who perceived the

target peer as similar to themselves were equally willing to participate in activities with

peers regardless of the presence or absence of emotional and behavioral problems.

Remarkably, the lack of differentiation in acceptance ratings (i.e., behavioral intentions)

among emotional and behavioral problems was largely accounted for by the large

increase in females' willingness to engage in activities with peers exhibiting conduct-

disordered behaviors if they had prior counseling experience themselves or perceived

themselves as similar to the target peer. Surprisingly, males did not make the same

differentiation or exception for their same-sex conduct-disordered peers. They continued

to be least willing to socialize with same-sex peers with conduct disorder relative to other









conditions, regardless of previous counseling experience and perceived similarity. On the

other hand, males' willingness to interact with peers with depression increased

significantly if they had prior counseling experience themselves. It is interesting that the

largest increases in willingness to participate in activities related to prior counseling

experience or perceived similarity were for females considering same-sex peers with

conduct disorder and males considering same-sex peers with depression-both disorders

that are seen most negatively by each gender, respectively.

These findings make intuitive sense, in that perceived similarity likely engenders

empathy and understanding, leading to greater tolerance and acceptance. Further support

for this explanation derives from the fact that only a small number of adolescents rated

themselves as similar to the target peer, particularly with respect to the target peers

described as having emotional and behavioral problems. Given the ambiguity of the

vignettes, it is likely that adolescents who may have experienced similar problems may

have felt more affinity (similarity) with these peers based on the characteristics of the

emotional and behavioral problems presented. Likewise, adolescents with prior

counseling experience undoubtedly experienced some symptoms of emotional and

behavioral problems themselves, and therefore have more empathy for peer with similar

problems.

Along these lines, adolescents' increased ability to take perspective in

combination with prior counseling experience or perceived similarity likely leads to

better understanding of factors contributing to emotional and behavioral problems and

generates more sympathy, or at least, less negative attitudes. Perhaps having prior

counseling experience or perceiving oneself as similar to a peer with emotional and









behavioral problems could provide validity for the experience of conduct disorder for

females and depression for males, problems not typically viewed as normative for each

gender, respectively.

Ethnicity also produced small moderating effects as a function of emotional and

behavioral problems. Specifically, African-American youth were significantly less

willing to socialize with peers exhibiting conduct-disordered behaviors compared to peers

with depression, family conflict, or no problems at all. This finding seems somewhat

surprising. African-American youth are stereotypically portrayed by the media as

endorsing or engaging in aggressive and delinquent behavior. Perhaps, because of this

negative stereotype, the African American community pays greater attention to

preventing youth from engaging in conduct-disordered behaviors and associating with

deviant peers. Caucasian youth were actually more willing to participate in activities with

peers who exhibited signs of depression relative to those with no problems. It is possible

that behaviors associated with depression are less problematic in social interactions for

Caucasian adolescents as compared to African-American adolescents; the exact reason

for this is unclear. Interpretations of these data should be guarded until these results are

replicated.

Individual characteristics also moderated several effects associated with

counseling status, but not necessarily as expected, and the effects were related to general

attitudes rather than behavioral intentions. For example, adolescents who perceived

themselves as similar to the target peers generally rated them more favorably, with the

exception of peers described as depressed and attending counseling. Contrary to

expectations, depressed peers who participated in counseling were rated less favorably if









adolescents perceived the target peer as similar to them. It is possible that peers who

perceived themselves as similar to the target peer in the depressed condition may have

been experiencing feelings of depression themselves. The ambiguity of the vignette

descriptions may have permitted projection of attitudes toward themselves. The lower

rating of peers with depression who attended counseling may reflect feelings of

hopelessness on the part of the respondent or feelings that depression is a problem that is

either not appropriate for or amenable to counseling. Likewise, if adolescents perceive

themselves as similar to the target peer with depression, they may be experiencing low

self-esteem, less optimism, or depressive realism and project those feelings onto the peers

in the vignette with similar problems. As a result, they attribute more negative descriptors

to them. Alternatively, perhaps an adolescent experiencing depression and not receiving

counseling may feel that depression is not a problem in need of counseling.

Consequently, they may have more negative attitudes toward peers with depression who

do attend counseling.

Slight grade effects also emerged, with males in the 12th grade rating their same-

sex peers lower if they participated in counseling relative to those who did not participate

in counseling. However, this unexpected and odd finding is contrary to previous literature

and theoretical explanations related to age differences. Consequently, interpretation of

this finding should be guarded until further research can either confirm or refute it.

Finally, and perhaps most notably, youths living in the country had less favorable

views of adolescents in counseling as compared to youth living in town; it is noteworthy

that findings related to rurality emerged, despite the somewhat imprecise designation of

rurality (i.e., asking adolescents whether they lived "in town" or "in the country").









Additionally, no interaction between rural living conditions and type of emotional and

behavioral problems emerged, suggesting that stigma for counseling was more evident

among youth from predominantly rural, less populated areas. For this group of

individuals, the stigma related to counseling may be more significant than stigma related

to emotional and behavioral problems displayed by individuals. Although perceived

stigma related to mental health services has been widely documented and associated with

rural areas and linked with reluctance to seek mental health services when available

(Berry & Davis, 1978; Stefl & Prosperi, 1985), this study empirically documented the

existence of stigmatizing attitudes within a rural community. This finding is not

necessarily surprising, however, given the potentially limited knowledge of rural

residents about the nature of mental health services, lack of knowledge about common

emotional and behavioral problems that counseling can effectively address, and

traditional values that are incompatible with help-seeking (Hoyt et al., 1997; Wagenfeld

et al., 1994).


Strengths and Future Directions

This study contributes significantly to the current literature on mental health

stigma among adolescents. Specifically, this study assessed stigma mental health

problems and treatment within the community setting rather than individually in the form

of perceived stigma. Although perceived stigma can be an important barrier to seeking

services, understanding the nature of attitudes prevalent in the community can facilitate

addressing this issue Do negative attitudes really exist? The results of this study suggest

that they do, but with qualifications, depending on the type of emotional and behavioral

problem displayed and personal characteristics, as described above.









This research also provides data on attitudes related to more common emotional

and behavioral problems rather than chronic, severe mental illnesses. Stigma is unlikely

to prevent individuals with severe mental illness for seeking or receiving treatment;

whereas stigma most certainly can impact individuals' willingness to seek help for less

severe, but nonetheless impairing, mental health problems. As such, stigma attached to

common emotional and behavioral problems is highly relevant. Similarly, this study goes

beyond the study of stigma attached to labels of "mental illness" by examining attitudes

toward behaviors associated with common emotional and behavioral problems. There is a

considerable body of literature suggesting that terms such as "mental illness" or "mental

disorder" generate negative evaluations, but much less is known about specific mental

health problems (with the exception of schizophrenia). This study also examines attitudes

within an adolescent population, which has been largely ignored in this literature, and

used developmentally appropriate, validated measures to empirically examine attitudes

rather than relying on unstandardized qualitative or descriptive information.

This study also advances the study of stigma, by operationally defining stigma as

an attitude and measuring both the evaluative and behavioral components of attitudes.

Similar to previous attitudinal research, the results of this study suggest that these

components of attitudes are distinctly different and yield different results. Consequently,

researchers should continue to assess multiple aspects of attitudes in order to develop a

more complete understanding of adolescent perceptions of and reactions to mental health

problems and counseling. Further, it is likely that the components of attitudes may relate

differently to actual behaviors.









A natural extension of this research would be to compare attitudes at the level of

the community with both perceived stigma at the individual level as well as actual

incidents of stigmatizing behaviors encountered by a clinic-referred, adolescent

population. One could also assess the role of the peer group and friendships as a source of

support for adolescents who have emotional and behavioral problems and seek help.

The vignette methodology used in this study also allows direct comparisons of the

stigma related to mental health problems as compared to and in combination with

counseling. The stigma associated with mental health problems and their treatment is

easily confounded. This is one of the few studies (cf., Philips, 1963) that attempted to

tease apart these attitudes, and it suggests that while negative attitudes are associated with

seeking counseling, they tend to be less prominent than those associated with emotional

and behavioral problems. Similar vignette paradigms could also be used to examine

differences in mental health attitudes as a function of location (rural vs. urban) and

service delivery (school vs. clinic).


Limitations

Although this study provides promising data about the prevalence of stigmatizing

attitudes associated with mental health issues, several limitations restrict its

generalizability. First and foremost, the study was hypothetical in nature, and as such,

interactions with peers and attitudes in real-life situations may differ. Although designed

to approximate typical behaviors, the stimuli presented were artificial and standardized in

order to provide experimental control. Adolescents evaluated the peers in the vignettes

based on behaviors exhibited in isolation. However, adolescent interactions with their

peers are certainly not determined in isolation and are likely colored by a variety of









factors in addition to the presence of emotional and behavioral problems or attending

counseling. Likewise, social desirability may have impacted responses, and more

negativity surrounding emotional and behavioral problems and counseling may be

present than indicated in this study. For example, adolescents could have answered the

questionnaires in a socially desirable or somewhat naive manner, suggesting that they

would not discriminate against peers with characteristics of emotional and behavioral

problems or who attend counseling. However, since the measure of social desirability did

not correlate with the dependent measures, it is likely that social desirability was not a

strong influence in responding.

Since adolescents only rated one target peer, it is possible that attitudes could

differ within each individual. While adolescents as a group made distinctions between

emotional and behavioral problems and counseling in this study, a single person may not

make such distinctions. Finally, although the effects found in this study were statistically

significant, their clinical significance is relatively unknown, particularly in terms of how

these attitudes correlate with real-life reactions to peers with emotional and behavioral

problems or who attend counseling.

It is also important to remember that the findings represent group averages. An

individual or certain factions of a peer group may hold very different attitudes than the

peer group as a whole. Unfortunately, there will likely be at least one person who

stigmatizes, or acts negatively toward, another; this can be a highly salient event

regardless of general peer group attitudes. A single individual who gossips, teases, or

excludes a person from activities, or at least the fear of being stigmatized by such an

individual, may be enough to prevent someone from seeking services. This is the