Clients' perceived trust, trustworthiness, and expectations in the therapeutic relationship

MISSING IMAGE

Material Information

Title:
Clients' perceived trust, trustworthiness, and expectations in the therapeutic relationship factors in satisfaction with counseling received
Physical Description:
xii, 169 leaves : ; 29 cm.
Language:
English
Creator:
Caesar, Celia Andersen, 1964-
Publication Date:

Subjects

Subjects / Keywords:
Psychology thesis, Ph. D
Dissertations, Academic -- Psychology -- UF
Genre:
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1996.
Bibliography:
Includes bibliographical references (leaves 159-168).
Statement of Responsibility:
by Celia Andersen Caesar.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 023780125
oclc - 35751951
System ID:
AA00013577:00001


This item is only available as the following downloads:


Full Text












CLIENTS' PERCEIVED TRUST, TRUSTWORTHINESS, AND EXPECTATIONS
IN THE THERAPEUTIC RELATIONSHIP: FACTORS IN SATISFACTION
WITH COUNSELING RECEIVED














BY


CELIA ANDERSEN CAESAR















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

1996



UrI ERSIT OF FLORICA LIBRARIES































To JOHNNY,

the wind beneath my wings















ACKNOWLEDGEMENTS


First of all, I would like to thank and acknowledge the

Creator through Whom all things are possible.

I extend a warm appreciation to my committee, Dr.

Carolyn Tucker, Dr. Ronald Akers, Dr. Shae Kosch, Dr.

Jaquelyn Resnick, and Dr. Robert Ziller, for their patience,

guidance, and support. I especially thank Dr. Tucker for her

unfailing belief that I would succeed in this endeavor, and

for her uniqueness in combining teaching, consultation, and

guidance with warmth, humor, and genuine concern for me

throughout my graduate career.

Acknowledgements are due to all those who in some way

contributed to this study. The University of Florida

Counseling Center's Research Committee, and former Director

Dr. James Archer, are affectionately thanked for allowing the

study to be conducted at the Center. Special thanks are

extended to the counselors and staff who implemented the

study, and without whom data collection would not be

possible. Likewise, the Director of Student Mental Health

Services, Dr. Jackie Ayers, and the counselors there are

gratefully acknowledged for their support of the study. To

the clients who graciously accepted to take part in the

study, I sincerely extend my appreciation for their


iii









willingness to put down in writing their feelings about the

counseling process.

I sincerely thank the students who volunteered to help

with the scoring of the data, and my friend, Deborah Brady

for helping with the verification of the data. I am deeply

grateful to Dr. John Dixon for his time, effort, and

expertise in the statistical analyses of the data.

I extend my heartfelt love and appreciation to my

parents Carmen Lydia Andersen and Marcus Criminatus Andersen

for their undying support and encouragement in my academic

pursuits. I warmly acknowledge my sister and brothers whose

love and support keep me grounded. I sincerely thank my

large extended family of aunts, uncles, cousins, and in-laws

who in one way or another have been impactful in my life. I

affectionately thank the family at St. Augustine's Catholic

Student Center, and especially the 9:30, choir for being my

spiritual support.

In a special way, I honor my grandparents: Gregorio

Belardo, Vicenta O'Neill de Belardo, Natividad Monell de

Andersen, and Henry Andersen. Theirs are the shoulders upon

which I stand.

A ray of sunshine entered my life three years ago, my

son Andres Malcolm. He is acknowledged for his uncomplicated

way of experiencing life, which keeps me centered on the

truly important things. The newest addition to our family is

Camille Breana, whose pure beauty brightens every day.

Finally, all my love to my husband, my friend, Johnny R.









Caesar, for everything: love, support, understanding,

patience, encouragement, and taking care of Andres and

Camille. I would certainly not have been able to complete

this dissertation without him by my side.
















TABLE OF CONTENTS


ACKNOWLEDGEMENTS ............... .. ...................... iii

LIST OF TABLES ........................................... viii

ABSTRACT ................... ................................. x

CHAPTERS

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

Statement of the Problem ...............................1
Social Influence Theory ............................. 1
The Therapeutic Relationship ........................4
Purpose of the Study ................................... 7
Clients' Expectations About Counseling ..............8
Trust/Trustworthiness ............................... 9
Clients' Satisfaction with Counseling ..............10
Significance of the Study ............................. 10

2 LITERATURE REVIEW .................................... 13

Expectations About Counseling .........................13
Trust in Counseling................................... 22
Premature Termination ................................. 33
Satisfaction with Counseling ..........................40

3 METHODOLOGY ........................................... 46

Subjects .............................................. 46
Instruments .............................................47
Procedure ............................................. 51
Hypotheses .......... ...................................56

4 RESULTS ................................................59

Preliminary Analyses .................................. 61
The Repeated Measures Analysis of Variance with
Counselor Trustworthiness as the Dependent
Variable ......................................... 63
The Repeated Measures Analysis of Variance with
Counseling Evaluation as the Dependent
Variable ......................................... 65
Tests of the Hypotheses ............................... 66










5 DISCUSSION................. ................... ..... 109

Summary of the Results ................................109
Discussion of the Results ............................117
Limitations of the Study ..............................129
Suggestions for Future Research....................... 131
Conclusion ........................................... 135
Implications .................................... ...... 137

APPENDICES

A EXPECTATIONS ABOUT COUNSELING (BRIEF FORM) ...........139

B LIKERT TRUST SCALE .................................. 140

C COUNSELING EVALUATION INVENTORY .......................141

D CLIENT SATISFACTION QUESTIONNAIRE .................... 145

E COUNSELOR SUMMARY SHEET ............................... 147

F CLIENT SUMMARY SHEET ................................. 148

G COVER LETTER (FIRST ADMINISTRATION) .................. 150

H INFORMED CONSENT FORM ............................... 151

I MEMORANDUM TO/FROM COUNSELORS .........................154

J COVER LETTER (SECOND ADMINISTRATION) ................. 155

K REMINDER LETTER ............ .......... ............. 156

L COVER LETTER (THIRD ADMINISTRATION) .................. 157

REFERENCES .. ................. ............................ 159

BIOGRAPHICAL SKETCH ..................... .................. 169
















LIST OF TABLES


Table page

4-1 Simple Statistics for the Expectation Factors........ 70

4-2 Simple Statistics for the Trust Variables ............ 72

4-3 Pearson Correlation Coefficients Among the Expectation
Factors and the Counseling/Counselor
Satisfaction Measures ............................. 74

4-4 Multiple Regression for Counseling Evaluation at Time2
with the Expectation Factors at Time2 as
Predictors ........................................ 76

4-5 Multiple Regression for Counseling Evaluation at Time3
with the Expectation Factors at Time3 as
Predictors ........................................ 78

4-6 Pearson Correlation Coefficients Among the Trust
Variables and the Counseling/Counselor
Satisfaction Measures ............................. 81

4-7 Multiple Regression for Client Satisfaction at Time2
with the Trust Variables at Time2 as Predictors ... 83

4-8 Multiple Regression for Client Satisfaction at Time3
with the Trust Variables at Time3 as Predictors ... 84

4-9 Multiple Regression for Counseling Evaluation at Time2
with the Trust Variables at Time2 as Predictors ... 85

4-10 Multiple Regression for Counseling Evaluation at Time3
with the Trust Variables at Time3 as Predictors ... 87

4-11 A Comparison of the Differences Between the Counselors'
and Clients' Perspectives of the Clients'
Termination Status ................................ 90

4-12 Analysis of Variance for Client Satisfaction at Time2
Based on the Counselors' Perspectives of the
Clients' Termination Status ....................... 95


viii









4-13 Simple Statistics Comparing Mutual Terminators and
Dropouts on the Client Satisfaction and
Counseling Evaluation Measures .................... 95

4-14 Analysis of Variance for Client Satisfaction at Time3
Based on the Counselors' Perspectives of the
Clients' Termination Status ....................... 96

4-15 Analysis of Covariance for Counseling Evaluation at
Time3 Based on the Counselors' Perspectives of
the Clients' Termination Status ................... 98

4-16 Multiple Regression for Client Satisfaction at Time3
Using Predictors from Timel ....................... 100

4-17 Multiple Regression for Client Satisfaction at Time3
Using Predictors from Time2 ....................... 102

4-18 Multiple Regression for Counseling Evaluation at Time3
Using Predictors from Timel ....................... 104

4-19 Multiple Regression for Counseling Evaluation at Time3
Using Predictors from Time2 ....................... 105

4-20 Perceived Client Similarity Ratings to Counselors for
Mutual Terminators and Dropouts ................... 108
















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


CLIENTS' PERCEIVED TRUST, TRUSTWORTHINESS, AND EXPECTATIONS
IN THE THERAPEUTIC RELATIONSHIP: FACTORS IN
SATISFACTION WITH COUNSELING RECEIVED


By

Celia Andersen Caesar

August, 1996


Chairperson: Carolyn M. Tucker
Major Department: Psychology

This study examined clients' expectations about

counseling, and perceived trust/trustworthiness before

receiving counseling. Changes in these variables and in

clients' satisfaction with counseling/counselors over time in

counseling were investigated. Also investigated were

differences between mutual terminators and dropouts regarding

the variables of study. Clients' perceptions of similarity

to their counselors were examined relative to clients'

satisfaction in counseling.

Seventy-seven clients seeking counseling at two

university counseling centers were administered the

Expectations About Counseling-Brief Form and the Likert Trust

Scale before counseling. These questionnaires were

readministered after the clients' first and sixth/last









counseling sessions along with the Client Satisfaction

Questionnaire and Counseling Evaluation Inventory.

The first four hypotheses stated that clients'

expectations, and perceptions of trust/trustworthiness, would

increase over the three specified time periods. T-tests and

repeated measures analyses of variance and covariance

revealed no significant changes in these variables over time.

Correlations indicated that number of counseling sessions

attended was not significantly related to clients'

expectations, or perceptions of trust/trustworthiness.

Hypotheses five and six combined stated that clients'

expectations and perceptions of trust/trustworthiness would

be significantly associated with clients' satisfaction with

counseling. Multiple regressions indicated that neither of

these variables were significant predictors of clients'

satisfaction with counseling.

Hypothesis seven stated that compared to dropouts,

mutual terminators would evidence higher expectations,

perceptions of trust/trustworthiness, and satisfaction with

counseling. Multivariate and univariate analyses of variance

and covariance revealed no significant differences between

the two groups on their expectations or perceptions of

trust/trustworthiness. However, based on the counselors'

designations of the clients' termination status, mutual

terminators were more satisfied with counseling than

dropouts.









Hypothesis eight stated that compared with precounseling

expectations and perceptions of trust/trustworthiness,

expectations, trust, and satisfaction early in counseling

would be better predictors of satisfaction later in

counseling. Regressions indicated that expectations, trust,

and satisfaction early in counseling accounted for sixty-

seven percent of the total variance in clients' satisfaction

with their counselors later in counseling.

Lastly, hypothesis nine stated that there would be

significant positive associations between clients' perceived

similarity to their counselors, and clients' satisfaction

with their counseling/counselors. Correlations supported this

hypothesis. Considerations for future research are

discussed.















CHAPTER 1
INTRODUCTION


Statement of the Problem


The goal of counseling is to help people change (Heppner

& Claiborn, 1989). Questions as to how to facilitate this

change is at the heart of the work of counseling

practitioners. Some therapists believe that client change

occurs naturally as a result of the therapeutic relationship

itself. Others maintain that a good relationship is

important inasmuch as it allows the counselor to exert

influence and use certain techniques to bring about change in

the clients' attitudes and behaviors. Regardless of their

therapeutic style, practitioners and theoreticians "probably

all currently agree that the counselor client relationship is

important to the outcome of most or all therapeutic efforts"

(Gelso & Carter, 1985, p. 155).


Social Influence Theory


Social influence theory offers an explanation of the

change process in counseling. This theory suggests that

there are two stages involved in this change process. First,

the counselor needs to establish herself/himself as a useful

resource. Second, after this perception has been









established, the counselor has the leverage needed to

influence clients therapeutically. Social influence theory,

therefore, conceptualizes counseling as an interpersonal

influence process. This process is defined as the ability of

one person to influence the actions, attitudes, or feelings

of another.

The social influence theory grew out of the social

psychology literature. Writers, such as Levy (1963), and

Goldstein, Heller, and Sechrest (1966), advocated a

connection between social psychological and psychotherapeutic

research. However, it was Strong's (1968) article that

introduced social influence as a viable theory of the

counseling process, and spearheaded research in counseling

based on the social influence theory.

In addition to proposing a two-stage model of the

counseling process, social influence theory also delineates

specific counselor variables that contribute to the

counselor's power to influence change in counseling. These

variables are perceived counselor expertness, attractiveness

and trustworthiness. A substantial amount of research has

been conducted to determine the effect of a range of

counselor-related variables (e.g., objective evidence of the

counselor's training, counselor verbal and nonverbal

behaviors, counselor attire, and office decor) on clients'

perceptions of their counselors' expertness, attractiveness,

and trustworthiness. As a tribute to the proliferation of

research in this area, three reviews of the social influence









literature have been published; two at the end of only the

first decade of social influence research (Corrigan, Dell,

Lewis, & Schmidt, 1980; Heppner & Claiborn, 1989; Heppner &

Dixon, 1981).

Social influence, therefore, has emerged as a major

research area in the counseling psychology literature within

the last twenty years. However, Heppner and Claiborn (1989)

have articulated several criticisms of this research First,

these researchers asserted that of the 56 empirical

investigations reviewed about social influence by them, 45

(80%) continued to examine the effect of certain counselor

behaviors and cues on perceptions of counselor expertness,

attractiveness, and trustworthiness. These studies,

therefore, examined only the first stage of Strong's (1968)

model. Most of the research has tended to ignore the

influence process, Strong's (1968) second stage.

Second, Heppner and Claiborn (1989) also asserted that

not enough attention has been given to client individual

difference variables in the social influence process. The

social influence research has historically treated the client

as a passive recipient of information. Heppner and Claiborn,

(1989) recommended that the client be conceptualized as an

active processor of information within the influence process.

As such, client characteristics should become more prominent

variables in the change process.

Finally, a third major criticism of social influence

research raised by Heppner and Claiborn (1989) is that the









majority of research in this area has been based on analogue

studies. These researchers urged movement away from the

overuse of brief analogue methodologies. They see as

absolutely essential the need for the social influence

process to be examined in realistic counseling situations.

In order to determine the research variables that would

be important in extending the knowledge base of the influence

process in counseling, an examination of the therapeutic

relationship was undertaken. The client-counselor

relationship is the vehicle through which counselors have the

ability to exert influence in promoting client change.

Therefore, it seemed essential to consider the determinants

of the therapeutic relationship.


The Therapeutic Relationship


It was from the client-centered therapy tradition that

much of the research about the therapeutic relationship

emanated. The late 1960s and early 1970s saw an abundance of

research and theory stemming from client-centered therapy.

The focus of this research was on the facilitative conditions

as being both necessary and sufficient for positive

counseling outcomes. However, more recently, research in

this area has diminished. Gelso and Carter (1985) suggested

that there is a natural cycle in the life of scientific

paradigms and that research on the facilitative conditions

has obtained its natural limit. Furthermore, Gelso and

Carter (1985) proposed that "what is now needed is a new









paradigm, a new way of thinking about the relationship in

counseling and psychotherapy" (p. 157). Consistent with the

idea of a new paradigm, an effort was made to empirically

define the therapeutic relationship.

It was Greenson (1967) who first suggested that the

therapeutic relationship could be divided into three

components: 1) the working alliance, 2) the transference

relationship, and 3) the real relationship. His work had a

major impact on analysis and analytic therapy. However, it

was Gelso and Carter (1985) who proposed that these three

components are not specific to psychoanalysis, but rather

exist in all therapeutic relationships. Moreover, the

theoretical perspectives and practices of the therapist

determines the importance a given component plays in therapy.

The working alliance is defined as the "alignment that

occurs between counselor and client...(based on) an emotional

bond between the participants, an agreement about the goals

of counseling, and agreement about the tasks of the work"

(Gelso & Carter, 1985). In order for a strong positive

working alliance to be formed, the client must have the

capacity to trust. As an essential part of the therapeutic

relationship, trust must be established between counselor and

client before any therapeutic changes can occur.

Furthermore, trust needs to develop early in the relationship

otherwise clients may terminate counseling prematurely or

remain resistant, sharing only superficial issues with the

counselor (Fong & Cox, 1983).









The transference relationship, also defined as the

"unreal" relationship, is a core construct in psychoanalytic

theory and as such was originally defined in terms of Oedipal

issues. However, as counseling psychologists began to

explore the potential value of this construct, it became

evident that a redefinition was needed. Consequently, Gelso

and Carter (1985) proposed the following definition: "a

repetition of past conflicts (usually but not always

beginning in early childhood) with significant others such

that feelings, behaviors, and attitudes belonging rightfully

in those early relationships are displaced; in therapy, the

displacement is onto the therapist." These misperceptions or

misinterpretations of the therapist may be positive or

negative and are only a part of the total relationship.

Gelso and Carter (1985) further suggest, that transference

occurs from the moment of first contact with the helping

person, and often even prior to that, in terms of

anticipations about the intervention and the helping person.

They label this type of transference as preformed

transferences or transference "expectancies" (Gelso & Carter,

1985). It would appear, therefore, that as clients come in

to counseling, they already have preformed ideas about their

counselor and the counseling process. Regardless, of where

these ideas find their origin or whether or not they are

within the client's conscious awareness, they greatly impact

on the therapeutic relationship.









Whereas transference is a core issue in psychoanalysis,

the real relationship is at the heart of humanistic

therapies. Unlike the transference relationship, the real

relationship consists of the counseling participants'

realistic perceptions of and reactions to each other. From

the moment of their first encounter, both parties actively

contribute to the real relationship. It exists alongside and

enmeshed with the transference relationship and is a part of

all therapies. The expectations counselors and clients have

for the real relationship regarding their respective roles

are crucial to successful therapy. If these expectations are

too discrepant, progress in therapy may be jeopardized and

the client may terminate therapy prematurely (Gelso & Carter,

1985).


Purpose of the Study


The purpose of this study, therefore, was to contribute

needed information to social influence theory with regard to

the change process in counseling. A review of the social

influence theory research suggested areas in need of further

investigation. Presented', is an examination of the basic

components of the therapeutic relationship, including

variables that impact the formation of a good client-

counselor relationship. These variables are clients'

expectations about counseling, perceptions of

trust/trustworthiness, and satisfaction with counseling.









Clients' Expectations About Counseling


It has been noted that clients' expectations of the

counseling process and of their counselor both impact on all

three components of the therapeutic relationship: the

working alliance, the transference relationship, and the real

relationship. These client expectations influence the

transference relationship, which as mentioned, exists prior

to and in anticipation of therapy. In turn, the transference

relationship impacts on the working alliance. In addition,

Gelso and Carter (1985) suggested that the working alliance

is the most basic of the components since it contains the

essential reasons for individuals to seek therapy to begin

with, namely their expectations to receive help. The real

relationship is also influenced by expectations in that these

expectations define the roles that the counselor and client

will play in the therapeutic relationship.

The interpersonal influence literature has a dearth of

research investigating the effect of client characteristics

on the interpersonal influence process. Clients'

expectations about counseling is one client variable that has

received substantial attention in the counseling literature

and it has recently found its way into the social influence

literature (Heppner & Heesacker, 1983). Not only is client

expectations an important client variable to be considered in

social influence research, in addition, Heppner and Heesacker

(1983) suggest that research investigate the predictive







9

utility of clients' expectations about counseling on clients'

satisfaction with counseling received (a potential measure of

counselor influence).


Trust/Trustworthiness


Of crucial importance to the working alliance and the

real relationship is the client's ability to trust and their

perception of the counselor as trustworthy. Although

psychoanalytic theory does not place much emphasis on the

working alliance or the real relationship, it is evident that

in all therapies it is essential for clients to believe that

their counselor will not mislead or injure them in any way.

The inability to form a trusting relationship impedes the

progress of therapy since it is doubtful that a working

alliance and a positive real relationship would be

established.

Trustworthiness has been the least frequently examined

variable in the social influence research. This lack of

research on trustworthiness as a component of the influence

process has been criticized as a weakness in the research

literature (Fong & Cox, 1983). Moreover, the research that

has been conducted has focused on trustworthiness as a

counselor characteristic. Social influence research has not

considered trust as a client variable although, as noted in

the therapeutic relationship literature, a client's ability

to trust is crucial in forming a working alliance with

his/her therapist.










Clients' Satisfaction with Counseling


Finally, clients' satisfaction with the counseling

process in early sessions has been shown to be one of the

best predictors of later satisfaction and other outcome

measures (Gomes-Schwartz, 1978). This finding suggests that

client satisfaction with counseling is a key ingredient in

the change process in psychotherapy. Specifically, it

impacts on the quality of the real relationship and is

essential in forging a strong working alliance.

In social influence research, client satisfaction with

counseling has been considered a measure of counselor

influence. Heppner and Claiborn (1989) indicated that

although satisfaction is a global and ambiguous variable, it

is certainly an important outcome, especially from a consumer

advocacy position.

A final important consideration in counseling research

is the applicability of research findings to a wide variety

of clients. Many research studies are limited in that their

findings are specific to the clients who are able to complete

the study. Therefore it is usually not known how dropouts

perceived their counselor or how satisfied they were with

counseling.


Significance of the Study


The current study sought to be a contribution to social

influence theory research. As mentioned before, the majority








11

of previous research in this area has been based on analogue

studies. There is a paucity of real-life counseling studies

investigating the importance of clients' expectations,

clients' trust levels, and clients' satisfaction with

counseling as important variables in the interpersonal

influence process. Therefore, the current research focused

on the importance of these client variables and their impact

on the influence process within a real-life counseling

situation.

More specifically, this study investigated:

1) the relationship among clients' satisfaction with

counseling, expectations about counseling, and perceived

trust/trustworthiness with the clients' decisions to stay in

counseling or to drop out;

2) whether expectations about counseling and perceived

trust/trustworthiness levels change over time and if so,

whether these changes are related to clients' satisfaction

with counseling; and

3) whether expectations about counseling, perceived

trust/trustworthiness, and clients' satisfaction with initial

counseling sessions predict clients' satisfaction with

counseling during later sessions and/or at termination.

It was proposed that the findings from this study would

provide:

1) an impetus for counseling centers to assess their

success at meeting the needs of their clients by obtaining an








12

estimate of the clients' levels of satisfaction with the

services they receive; and

2) information that might stimulate strategies for

increasing the retention of clients in counseling.

Especially for novice clients, it may be important for

counselors to discuss the client's expectations for

counseling, especially the roles of the counselor and the

client, and to intermittently check with the client as to

their satisfaction with the progress of counseling.
















CHAPTER 2
LITERATURE REVIEW



This chapter focused on the relevant literature

pertaining to the research variables that were investigated

in this study. These variables were clients' expectations

about counseling, clients' perceptions of their own and their

counselors' trust/trustworthiness, clients' premature

termination from counseling, and clients' satisfaction with

counseling.


Expectations About Counseling


Clients' expectations about counseling has been an area

of great research interest in the field of counseling

psychology over the past ten years. It has been well

established that clients come to counseling with certain

expectations about what counseling will be like and what

roles they and their counselors will assume (Bordin, 1955;

Patterson, 1958; Rosenthal & Frank, 1956; Strong, 1968;

Tinsley & Harris, 1976). Furthermore, there has been support

for the argument that clients' expectations may either

facilitate or hinder the effectiveness of the counseling

process (Apfelbaum, 1958; Frank, 1968; Goldstein, 1962;

Goldstein, Heller, & Sechrest, 1966).









In 1979, Duckro, Beal, and George noted the failure of

many researchers to distinguish between expectations about

counseling and preferences for counseling. This lack of

distinction has caused many inconsistencies in the

literature. Tinsley, Bowman, and Ray (1988) later reaffirmed

this problem and noted that an additional confound in the

expectancy literature is the use of clients' perceptions of

counseling. In response to this issue Tinsley, et al. (1988)

suggested that the full range of client expectancies be

identified and that a reliable and valid measure of those

expectancies be developed. Tinsley, Workman and Kass (1980)

developed the Expectations About Counseling Questionnaire

(EAC) to address this need. Although this questionnaire has

not been published, it has been used in several studies

(e.g., Craig & Hennessy, 1989; Hardin, Subich, & Holvey,

1988; Heppner & Heesacker, 1983;) and has been shown to

measure constructs unique from those measured by perception

instruments (Hayes and Tinsley, 1989).

Much of the research on expectations about counseling

has focused on the demographic characteristics of

respondents. For example, studies have shown that women

expect that their counselors will be confrontational,

genuine, nurturing, tolerant, trustworthy (Subich, 1983), and

attractive, and expect that they themselves will be more

responsible and self-motivated in counseling than men (Hardin

& Yanico, 1983). However, men expect counselors to be

directive, critical, analytical (Tinsley & Harris, 1976), and










self-disclosing (Subich, 1983). Tinsley and Harris (1976)

found that the longer students remained in college, the less

they anticipated a counselor to be accepting, and an expert.

Richmond (1984) found that compared to clients, nonclients

had significantly higher expectations that counselors would

be genuine, attractive, and trustworthy.

With respect to the influence of cultural factors on

clients' expectations, Yuen and Tinsley (1981) found that

international students expected their counselor to be more

concrete, directive, empathic, nurturant, and expert than did

American students. American students reported a higher

expectancy than international students to take action and to

admit responsibility for counseling progress. Consistent

with this finding, Hector and Fray (1987) also reported that

Asian clients expected their counselor to be directive and

nurturing. Hispanic clients preferred their counselors to be

active and directive. In addition, the client's attitude

about the counselor and not the counselor's ethnic background

was the Hispanic client's most important consideration.

American Indians expected to discuss more school-related

concerns and had a strong preference for Native American

counselors. Compared to American students, African students

studying in the USA expected counselors to be more directive

and nurturing.

In a study by Proctor and Rosen (1981), clients with

definite preferences preferred counselors of their own race.

However, both White clients and Black clients expected their









counselor to be White. Although Blacks had a strong

preference for a counselor of the same race, failure to meet

this preference was not significantly related to premature

termination. In addition, neither groups' satisfaction with

counseling was significantly related to their expectation or

preference with regard to the counselor's race. However,

this study used only White counselors and it cannot be

determined if Blacks fared as well with a White counselor as

they would have with a Black counselor.

Recent studies have focused more on core personality,

attitudinal, and cognitive attributes. For instance Craig

and Hennessy (1989) used the Conceptual Systems Theory (CST)

as "a framework from which to assess the degree to which a

stable, preexisting personality dimension can explain the

variance in precounseling expectations (p. 402)." Conceptual

systems are defined as the predispositions that persons use

to construe and interpret perceptions. The CST holds that

there are four discrete stages of conceptual development.

Eleven subscales of the Expectations About Counseling

Questionnaire (EAC; Tinsley et al., 1980) were subjected to a

discriminant function analysis.

The two significant discriminant functions that

distinguished the four conceptual stages concerned

expectations that clients had about counselor

characteristics, attitudes and behaviors. More simply

stated, compared to each other, individuals in different

stages of conceptual development had different expectations







17

about counselor empathy, directiveness, attractiveness (first

function) and counselor self-disclosure, genuineness, and

nurturance (second function). Craig and Hennessy (1989)

suggested that the degree to which these expectations are

confirmed or refuted can have an impact on the initial

progress of counseling, and the counseling outcome.

A study by Tinsley, Hinson, Holt, and Tinsley (1990)

investigated the relation between students' levels of

psychosocial development and their expectations about

counseling. Results showed that students with more

appropriate educational plans, more mature career plans and

more mature lifestyle plans had more positive expectations

for counseling. The authors suggested that working with less

mature college students, counselors may need to focus early

in the counseling process on helping the students develop a

sense of personal responsibility for the success of

counseling.

Karzmark, Greenfield, and Cross (1983) investigated the

relationship between clients' levels of adjustment and

expectations for therapy. Subjects were 110 new clients at a

university counseling center. Clients estimated their level

of adjustment by completing a questionnaire consisting of 25

items selected from the SCL-90 symptom checklist (Derogatis,

Lepman, & Covi, 1973. Therapists assessed clients'

adjustment at the end of the intake session using the Global

Assessment Scale (GAS; Endicott, Spitzer, Fleiss, & Cohen,

1976). In addition, clients' expectations were assessed by a








18

5-item, factor analytically derived scale. Clients assessed

therapy outcome by completing the Client Satisfaction

Questionnaire (CSQ; Larsen, Attkisson, Hargreaves, & Nguyen,

1979), and therapists evaluated therapy outcome by

reassessing clients with the GAS at termination.

Therapists' assessments of the clients' adjustment with

the GAS were used to predict clients' expectations. The

regression performed revealed a significant but small

positive linear relationship. Clients' own evaluations of

adjustment were not significant predictors of clients'

expectations. Clients' expectations for counseling also bore

a small but significant positive linear relationship to their

satisfaction. Clients' expectations were not related to

therapists' ratings of outcome.

A study by Heppner and Heesacker (1983) examined client

satisfaction in relation to clients' expectations about

counseling and their perception of counselor characteristics.

Seventy-two clients at a university counseling center

completed the study. Clients' perceptions of their

counselors expertness, attractiveness and trustworthiness

were obtained by the Counselor Rating Form (CRF; Barak &

Lacrosse, 1975). Forty-five items, constituting six scales

of the EAC were used to measure clients' expectations of

their own openness and motivation in counseling, and their

expectations of their counselor's acceptance, expertness,

attractiveness, and trustworthiness. Clients' satisfaction

with counseling was measured by the Counseling Evaluation










Inventory (CEI; Linden, Stone, & Shertzer, 1965). Clients

completed the EAC before their initial session and completed

the CRF and CEI two weeks before the end of the semester.

Results indicated that in general, the clients perceived

their counselors as being expert, attractive, and

trustworthy. In addition, clients rated their satisfaction

positively. Clients' perceptions of their counselors'

expertness, attractiveness and trustworthiness were

significantly correlated to clients' satisfaction with

counseling. Moreover, clients' expectations of their

openness and counselors' trustworthiness were also

significantly correlated with clients' satisfaction and

perceived counselor attractiveness, expertness and

trustworthiness. Regression analyses indicated that the CRF

variables were the best predictor of client satisfaction.

However, the EAC variables (acceptance, openness, motivation,

trustworthiness, attractiveness, expertness) did not

significantly predict client satisfaction nor counselor

expertness, attractiveness and trustworthiness (CRF).

These findings suggest that specific precounseling

client expectations were not predictive of client

satisfaction or perceived counselor characteristics after

counseling. One reason for this finding may be that before

entering counseling, clients have little knowledge of what to

expect from counseling, therefore specific clients'

expectations of counseling are based on minimal information,

or are not well founded. As they experience counseling their










expectations may change. These later expectations may be

more reflective of actual therapy experience and thus be more

predictive of client satisfaction than expectations prior to

counseling (Heppner & Heesacker, 1983).

The authors suggested that additional research be

conducted to examine how and to what extent client

expectations change over the course of counseling. In

addition, subjects who terminated prematurely from the study

did not differ on initial EAC scores from subjects who

remained. However, it is unknown how these individuals

perceived their counselor or how satisfied they were with

their counseling. Again the authors suggested that research

in this area would extend the present findings by examining

premature terminators.

Tinsley et al. (1980) conducted a study that factor

analyzed client expectancies as measured by the EAC. The

factors obtained were Personal Commitment, Facilitative

Conditions, Counselor Expertise and Nurturance. The authors

believe that moderately high scores on these factors may be

more desirable than extremely high scores. Extremely high

scores on the Personal Commitment factor may indicate

compulsive, perfectionistic tendencies on the part of the

client, whereas low scores may characterize naive clients who

expect the counselor to cure them without any effort on their

part. Extremely high scores on the Facilitative Conditions

factor may indicate that the client expects the counselor to

act in some idealistic but unobtainable manner, whereas low








21

scores may indicate a defensive, guarded client who may avoid

any revelation of affect. High scores on the Counselor

Expertise factor may be indicative of magical thinking by

clients, whereas low scores may indicate fatalism or

pessimism regarding the counselor's ability to really help

the client. On the Nurturance factor, extremely high scores

may represent a desire to escape to a warm pain-free

environment and low scores may represent a client's

difficulty in believing that the counselor or anyone else

could actually care for him or her.

In summary, the expectation literature has contributed

valuable information directly related to the process and

outcome of counseling. Research has addressed demographic

variables relative to client expectations for counseling and

have proceeded to examine more core personality factors.

Researchers have suggested that there needs to be a clear

delineation between counseling expectations, and perceptions

and preferences in order to avoid ambiguities and

inconsistencies in the literature. In addition, Hayes and

Tinsley (1989) encourage the use of the EAC as a prognostic

indicator, as a measure of client resources and attitudes

that are brought into counseling, and as an indicator of

whether education about what to expect during counseling may

be necessary before therapy is initiated.

Tinsley et al., (1988) suggested that from expectancy

research should come an understanding of the conditions under

which expectancy change might be thought to have beneficial







22

results. Fellows (1985) proposed that measuring differences

in expectations across time may offer a useful method for

studying changes in expectations. Finally, a follow up study

by Tinsley and Benton (1978) to the Tinsley and Harris (1976)

study confirmed that many potential clients may never seek

counseling because of their low expectation that they will be

helped.

An implication of these studies is that counselors need

to explore clients' expectations for counseling. Torey

(1972) believes that Western psychotherapists can learn a

great deal from African witch doctors. He found that witch

doctors typically consider the expectations of their clients

and that these expectations are critical to the success of

the treatment. Richmond (1984) suggests that the initial

counseling relationship should reflect the needs of clients

in response to expectations and that clients with lower

expectations may require additional relationship-building

interactions to move effectively toward change.


Trust in Counseling


The major focus of research on trust has been the

client's perception of the counselor as a trustworthy person.

The interpersonal influence research has contributed much

information in this area. However, there has been very

little research related to clients' ability to trust as a

personality variable. Work in this area has predominantly

investigated demographic differences and differences between









high and low trusters. In relation to race, Terrell in

conjunction with other associates has contributed much needed

information relative to Blacks' mistrust of Whites, including

the development of a scale to measure cultural mistrust

(Terrell, & Terrell, 1981).

It has been well established that the development of

trust is crucial to the effectiveness of counseling (Fong &

Cox, 1983; Johnson & Noonan, 1972; Rogers, 1951; Williams,

1974). Corazzini (1977) describes trust as including several

concepts: expectancy, reliance upon others, faith,

surrendering of control, consistency, mutuality, and utility

for risk. Rotter (1971) defines trust as "an expectancy by

an individual or a group that the word, promise, verbal, or

written statement of another individual or group can be

relied on" (p. 444). In counseling, Fong and Cox (1983) note

that "trust is the client's perception and belief that the

counselor will not mislead or injure the client in any

way"...(and) "is one of the crucial issues in the first stage

of counseling" (p. 163).

A study by Lagace and Gassenheimer (1989) using the

MacDonald, Kessel, and Fuller Self-report Trust Scale (1972)

showed that there were no differences between men and women

on trust. However, differences were found on the Suspicion

scale, with men scoring as more suspicious.

Rotter has been influential in the area of interpersonal

trust with the development of his Interpersonal Trust Scale

(1967). Some demographic data obtained with this instrument









indicated that youngest children were less trusting than

only, oldest, or middle children. Also college students

identifying themselves as atheists or agnostics were

significantly less trusting than others. Students who

perceived their parents as believing in two different

religions were less trusting than those who perceived both

parents as believing in the same religion or lack of religion

(Rotter, 1967).

In addition, a study by Terrell and Barrett (1979)

investigating the differences in interpersonal trust between

lower and upper socioeconomic classes, men and women, and

blacks and whites found that males were more trusting than

females, higher socioeconomic group members were more

trusting than lower ones, and White students were more

trusting than Black students. In relation to gender

differences, a contrasting finding was reported in a study by

Kaplan (1973) in which factors were derived for Rotter's

Trust Scale (1967). Males demonstrated significantly less

trust than females on each of the trust factor dimensions.

Grace and Schill (1986) investigated the differences in

social support and coping styles of subjects high and low in

interpersonal trust. Students in an introductory psychology

class were administered instruments measuring their level of

interpersonal trust, social support during the past month,

perception of social support from friends and family, and

coping style. Results showed that subjects high in trust

viewed both friends and family as more supportive than










subjects low in trust. Subjects high in trust reported

seeking more social support when feeling stressed, and

displayed less dysfunctional behavior than subjects low in

trust. Subjects low in trust became dysfunctional under

stress. The authors concluded that "a pervasive expectancy

that people cannot be trusted very likely negatively colors

the interpretation of reassurance and assistance, and results

in a perception of people as basically unsupportive" (p.

585).

Williams (1974) conducted a study in which Black college

students attended five 60-minute counseling sessions with

either a white male professional counselor, or a black male

or female peer counselor. Subjects completed questionnaires

related to their level of interpersonal trust and self-

disclosure before and after counseling. Subjects'

precounseling data on the self-disclosure and trust

questionnaires were used as a selection criteria to

participate in the study.

Although there were no significant differences between

the two groups on trust and self-disclosure after counseling,

mean differences showed that students established higher

levels of self-disclosure with the peer counselors (I =

133.89; SD = 23.09) than with the professional counselors (M

= 122.67; SD = 23.10) after counseling. This finding was not

obtained with regards to trust (peer group: I = 18.22; SD =

2.35; professional group: M = 19.67; SD = 4.55). In

addition, differences between the professional counselors and








26

peer counselors were not significant. Moreover, both groups

established significantly higher levels of self-disclosure

and trust after counseling treatment.

These findings suggest that the belief that White

professional counselors are less effective than peer

facilitators in promoting self-disclosure and trust in Black

clients is questionable. In addition, Black trained peer

counselors can be effective facilitators of self-disclosure

and trust among Black clients.

In a study by Wright (1975), 65 White and 35 Black male

students were pretested on the Interpersonal Trust Scale and

designated as either high trusters or low trusters based on

one standard deviation above or below the mean of this base

population. Means and standard deviations for the two racial

groups were as follows: Black students M = 61.67, SD =

9.11; White students M = 68.5, SD = 9.69. From the 100

students tested, 24 were chosen as clients; six high trusters

and six low trusters for each race. These students were then

administered the Barrett-Lennard Relationship Inventory

(Barrett-Lennard, 1962) prior to their first counseling

session and at the end of the fifth counseling session. The

counselors were two White, and one Black male with advanced

degrees in counseling, and one Black male psychology

practicum student.

Results from the pretest indicated that Black high

trusters believed that Black counselors would significantly

show a higher level of regard 1 (affection, liking), regard 2









(respect, appreciation, valuation), congruence 1 (freedom

from threat and anxiety), unconditionality 1 (uncritical

acceptance), unconditionality 2 (degree of agreement between

persons in the relationship), empathy 1 (degree of

recognition of another's perceptions or feelings that are

directly communicated), and empathy 2 (ability to place the

immediate experiencing of another in the context of the

objects or persons who produce that experience), than White

counselors. Black low trusters also expected Black

counselors to significantly express higher levels of regard

2, congruence 1, and empathy 1 in the counseling relationship

than White counselors. White high and low trusters did not

have significantly different expectations of the counselors

of the two racial groups prior to counseling.

The pre- and post-ratings of both ethnic groups were

compared to determine if changes in students' perceptions

occurred over time after counseling. There was a significant

difference in counselor race in the congruence 2 (degree of

what is felt and experienced) and unconditionality 3 (degree

of variability in one's affective response to another). Both

races perceived White counselors as communicating a higher

degree of what is expressed in the counseling relationship

than Black counselors. In addition, scores showed a reverse

mean pattern for regard 1 between the races. White

counselors had a higher degree of liking and affection for

Black clients than White clients, and Black counselors

communicated a higher degree of liking and affection for









White clients than Black clients. Empathy 3 (ability to

adopt another's frame of reference) was highly significant

for student race, counselor race, and trust level. High

trusters of both races believed that counselors of both races

could adopt another's frame of reference. Low trusters of

both racial groups perceived White counselors as incapable of

adopting another person's frame of reference. White low

trusters also felt the Black counselors could not adopt

another's frame of reference. However, Black low trusters

believed that Black counselors could adopt another person's

frame of reference.

Results of this study indicate that Black clients and

White clients enter the counseling relationship with

preconceived feelings about the opposite race. Moreover, if

therapy focuses on the here-and now, race does not inhibit

counselors from respecting, uncritically accepting, and

recognizing the perceptions or feelings of a different race.

This is exemplified by the finding that regardless of

counselor race, clients of the opposite race experienced a

change in perceptions (negative to positive), relative to the

counselors ability to show affection and liking, after

counseling. A concern with this study is the question of

whether or not counselors were blind to the purpose of the

study. The author stated that "prior to counseling, the

counselors received the necessary details of the project" (p.

163). The extent of these details was not described.









In an analogue study by Watkins and Terrell (1988),

Black students mistrust level and counseling expectations

were investigated in relation to Black client-White counselor

relationships. In addition to a demographic data

questionnaire, students were administered the Cultural

Mistrust Inventory (CMI; Terrell & Terrell, 1981), indicating

the degree to which they distrust Whites, and the

Expectations About Counseling Questionnaire (EAC; Tinsley et

al., 1980) which was modified for this study. The word

"black" or "white" was placed before counselor in the

instructions that the experimental groups received. Subjects

mistrust levels were considered high or low based on a median

split.

Results indicated that compared to less mistrustful

Blacks, highly mistrustful Blacks expected less from

counseling, regardless of the race of the counselor.

Specifically, highly mistrustful subjects viewed the

counselor as less attractive and expected a diminished focus

on the immediate therapy relationship than less mistrustful

subjects. Moreover, highly mistrustful subjects assigned to

a White counselor expected their counselor to be less

genuine, self-disclosive, accepting, trustworthy, and expert,

and expected counseling to be less successful than if they

were assigned to a Black counselor.

In a follow-up study, Watkins, Terrell, Miller, and

Terrell (1989) once again examined cultural mistrust in

relation to Black client-White counselor relationships. In









this study, instead of the EAC, subjects completed the

Counselor Effectiveness Rating Scale (CERS; Atkinson &

Wampold, 1982) which measures counselor credibility.

Additionally, clients completed the Personal Problem

Inventory (PPI; Schneider, 1985), a 20-item personal problem

list. Subjects rated each problem area in relation to their

perceived confidence in the counselor's ability to help them

with it. A one item measure of clients' willingness to see

the counselor was also used. Responses were made on a 6-

point Likert scale ranging from very unwilling (1) to very

willing (6). Two manipulation checks were employed; one

indicated the degree to which the subjects could identify

with the client role and the other indicated the subjects'

ability to identify the race of the counselor he/she read

about in the experimental manipulations. The counselor

description was identical across experimental groups with the

exception of the words "black" and "white" being placed

alternatingly in front of the word counselor.

Results indicated that highly mistrustful Blacks tended

to view the counselor as less credible when the counselor was

White. In addition, highly mistrustful Blacks viewed the

counselor as less able to help them deal with sexual

functioning concerns, regardless of the counselor's race.

Moreover, Blacks who were highly mistrustful of Whites viewed

the White counselor as being less able to help them deal with

general anxiety, shyness, dating difficulties, and feelings

of inferiority than a Black counselor.









This study and the previous one described, emphasized

the need for counselors to be sensitive to the mistrust issue

and its influence on the Black client-White counselor

relationship. Watkins, et al., (1989) identified specific

ways in which mistrust levels impact Blacks' expectations of

White counselors. The authors have suggested that

sensitivity to the mistrust issues in conjunction with the

specific problem areas identified would be valuable

counseling information for counselors to be aware of. In

addition, future research would need to use actual clients

and assess the effects of mistrust on counseling over time to

address the limitations of the two studies.

Finally, a study by Thompson, Neville, Weathers, Poston,

and Atkinson (1990) investigated cultural mistrust and racism

reactions among African-American students. Eighty-seven

African-American and 70 Euro-American college students took

part in the study. They completed an instrument consisting

of a demographic section, the Racism Reaction Scale and the

Interpersonal Relations, and Education and Training subscales

of the Cultural Mistrust Inventory (CMI; Terrell & Terrell,

1981). The Racism Reaction Scale (RRS) consists of 19 items

that were racism reaction statements by African-American

students and 19 additional items from the California

Psychological Inventory, used as filler items. Students were

categorized as high or low on the CMI using a median-split

method.










This study found that the African-American college

students felt that they were singled out for differential and

inferior treatment more so than their Euro-American

counterparts. The feeling was strongest among highly

mistrustful African-Americans. Although items on the RRS

reflecting generalized reaction to the environment revealed

no significant differences between the two racial groups,

African-American students reported higher means on more

situational/classroom-related items than Euro-Americans.

Moreover, the item on the RRS with the greatest difference in

ratings by African-American respondents and Euro-American

respondents was "I have to be prepared to deal with a

threatening environment."

These findings indicate that counselors need to become

more aware of racism reactions and be capable of

distinguishing these types of statements from statements

reflecting paranoia. In addition, some African-American

students seeking counseling for issues related to racism

reactions may be distrustful of any Euro-American counselor

and may, at least initially, express a preference for an

African-American counselor. Furthermore, the fact that the

staffs of most counseling centers are predominantly or

entirely White (Ponterotto, Anderson, & Greiger, 1986)

underscores the need for counselors to be sensitive to

African-American students' concerns about racism and their

possible reactions to seeing a White counselor. Lack of

sensitivity to these issues, and the application of a "White-










American student standard" to self-disclosures by African-

American students, that may actually be reactions to racism,

may frustrate African-American students in counseling and

further reinforce their feelings of alienation from the

university in general and campus counseling services in

particular (Thompson, et al., 1990)


Premature Termination


Premature termination poses a major obstacle to the

delivery of counseling services. As many as 20% to 25% of

college counseling center clients "no-show" for their first

appointment (Epperson, Bushway, & Warman, 1983) and

approximately 30% to 60% of all outpatient psychotherapy

clients drop out of treatment prematurely (Pekarik, 1983).

Premature termination, in its broadest sense, refers to

clients leaving treatment before their counselors believe

they should. This may be immediately after intake or after

the client has met with his or her counselor for one or more

sessions.

For those individuals who do not show for their first

appointment, the problem may be that of having to wait for a

counselor to pick up the individual as a client. Cochran and

Stamler (1989) suggest that demographic or administrative

variables, such as the client's dissatisfaction with the sex

of the intake counselor, length of the intake interview,

number of days from intake to case assignment, or the

experience level of the assigned counselor, could be more










salient factors in the client's decision to not return after

the intake. Archer (1984) suggests that reciprocal referrals

to other agencies, augmenting services with paraprofessionals

and prepackaged counseling programs and using groups and

workshops in lieu of individual counseling could be used to

reduce waiting lists. For individuals who terminate after

one or more counseling sessions, factors related to the

counseling process might play a more significant role in

nonmutual terminations (Cochran & Stamler, 1989)

Premature termination is typically viewed as a negative

treatment outcome. However several authors have questioned

this belief noting that premature termination does not always

reflect a negative outcome from the client's perspective

(Cochran & Stamler, 1989; Gunzberger, Henggeler, & Watson,

1985; Kokotovic & Tracey, 1987). Clients may feel that their

problem has been solved, their symptoms have been

substantially reduced or they have developed adequate support

systems outside the counseling setting (Mennicke, Lent, &

Burgoyne, 1988).

Although there has been very little research

investigating the welfare of dropouts, one study showed that

for a sample of clients failing to show for their first

appointment after receiving an intake interview and being

placed on a waiting list, only 29% reported having resolved

their presenting problem and 50% had sought help elsewhere

(Christensen, Birk, & Sedlacek, 1977). To the extent that

premature termination does not represent a positive treatment










outcome, it is a viable concern in process and outcome

research. Mennicke et al. (1988) suggest that much

counseling research is being conducted with a biased sample,

because treatment dropouts are often not considered.

In a study investigating differences between mutual and

nonmutual terminators in a university counseling center

(Cochran & Stamler, 1989) clients were mailed a questionnaire

seeking information on the client's satisfaction with

counseling and reasons for terminating counseling. Only

clients who had returned for at least one counseling session

after the intake interview were included in the study.

Clients who terminated counseling prematurely were

significantly less satisfied with the counseling experience

than clients who had planned a termination with their

counselors. In addition, proportionately more mutual

terminators endorsed having met their goals for counseling,

and proportionately more nonmutual terminators did not think

their counselors had the skills to help them. Moreover, none

of the 24 nonmutual terminators had sought help elsewhere and

only 4 (16%) felt that their situation had improved on its

own.

Cochran and Stamler's (1989) study is consistent with a

previous study by McNeill, May, and Lee (1987) that

investigated perceptions of counselor source characteristics

by premature and successful terminators. Fifty-six premature

terminators and 148 successful terminators completed a 2-page

evaluation form including the Client Satisfaction










Questionnaire (CSQ; Larsen et al., 1979), the Counselor

Rating Form (CRF-S; Corrigan & Schmidt, 1983) and a number of

miscellaneous and open-ended items for assessing client

satisfaction with various aspects of the counseling center's

services. Students were given the option of signing the

completed evaluation forms. Premature terminators defined as

clients who initiated their own terminations of counseling

without the knowledge or against the advice of the assigned

counselor, attended an average of 3.04 sessions (range = 0 -

17; mode = 2.0). Successful terminators, defined as clients

who attended three or more sessions and terminated by mutual

agreement with their counselor, attended an average of 7.38

sessions (range = 1-16; mode = 10.0).

The university counseling center clients who terminated

prematurely were significantly less satisfied with counseling

services that they received. In addition, premature

terminators viewed their counselors as significantly less

expert, attractive and trustworthy than did clients who

terminated successfully. However, premature terminators who

continued longer in counseling expressed more satisfaction

regardless of the perceived expertness, attractiveness, and

trustworthiness of their counselors. Subjects who did not

sign their evaluation form viewed their counselors as

significantly less attractive and trustworthy and expressed

less satisfaction with services than those who chose to sign

their forms.










These two studies are at variance with a study by

Gunzberger et al. (1985) which studied factors related to

premature termination of counseling relationships.

Gunzberger, et al. found no differences in client-reported

counseling outcome between clients who mutually terminated

with their counselors and clients who initiated their own

termination, contacted four weeks after termination. The

difference between this study and the two previous studies,

however, is that Gunzberger, et al. used a telephone

interview as opposed to anonymous questionnaires to gather

client information. This procedure may have been more

intrusive and thereby render the findings as less valid. It

has been found that anonymity reduces the high ceiling

effects present in reports of client satisfaction or

evaluation (McNeill, et al., 1987; Soelling & Newell, 1984).

Terrell and Terrell (1984) studied premature termination

as a function of counselor race, client sex, and cultural

mistrust level. One hundred and thirty-five black clients at

a community mental health center completed the Cultural

Mistrust Inventory (CMI; Terrell & Terrell, 1981) before

counseling. Clients were then randomly assigned to either a

Black or White counselor for an intake interview and

continued counseling. Premature terminators were defined as

individuals not returning for the second or any subsequent

counseling sessions.

Results showed that of the 68 clients seen by a Black

counselor, 17 (25%) did not return for a second session. Of










the 67 clients seen by a White counselor, 29 (43%) did not

return for additional counseling. Moreover, hierarchical

regression analyses were used to examine the relation between

the predictor variables of counselor's race, mistrust level,

and clients' sex on the criterion variable of premature

termination for counseling. The main effect of counselor

race accounted for 12% of the variance, and when added, the

main effect of trust, accounted for a significant increase in

the amount of variance (R2 = .18). This finding indicates

that Black clients were more likely to terminate from

counseling prematurely when seen by a White counselor than

when seen by a Black counselor. The trust data indicated

that regardless of counselor race, highly mistrustful clients

terminated treatment prematurely. Moreover, a significant

interaction was found between counselor race and client

mistrust level. Highly mistrustful clients seen by a White

counselor had a higher rate of premature termination than

highly mistrustful clients seen by a Black counselor.

Hardin et al. (1988) examined the relationship between

premature termination and clients' expectations about

counseling. Data for this study were drawn from archival

files at a university counseling center. Half of the

selected files consisted of clients who had come for one

appointment and, despite having made a second appointment,

never returned to the center. The remaining cases consisted

of clients who pursued counseling for two or more sessions (M

= 6.35 sessions, range = 3 19) to a mutually agreed-upon










termination. An equal number of clients reported career

concerns and personal concerns. Prior to their first

session, clients completed the Expectations About Counseling

questionnaire (EAC; Tinsley et al., 1980).

Results showed that precounseling expectations did not

differentiate terminators from nonterminators. In addition,

problem type did not affect expectations. Clients in both

groups, had high expectations about the counseling and

expected to assume a high degree of responsibility for their

counseling experience. Moreover, all clients expected to be

open with the counselor and to be at least moderately

motivated. They also had moderate levels of expectation for

counselor attractiveness, expertness, trustworthiness,

tolerance, nurturance, genuineness, acceptance, and

confrontation. They had lower expectations for counselor

directiveness, empathy, and self-disclosure.

The authors suggested that accepting unequivocally that

precounseling expectations do not affect termination status

is probably premature, sighting that one confounding factor

may result from clients who make subsequent appointments in

order to conform to their counselor's desires although they

themselves may feel satisfied after one interview. In

addition, it may be unwarranted to assume that clients

automatically distinguish between preferences and

expectations when asked to express their expectations. This

confusion between the two variables could have also

influenced the results of the study. Finally, Hardin et al










(1988) suggests that the EAC might be a measure of a global

positive or negative set toward counseling rather than a

number of discrete expectations. This study investigated

differences on each of the 17 subscales of the EAC, wherein

scales may be composed of only three or four items. Factor

scores may have proven more meaningful.

In a review article by Mennicke et al. (1988), it was

noted that studies investigating client satisfaction with

services and subsequent attendance has resulted in mixed

findings. Several researchers have failed to find a

relationship between client satisfaction with counseling and

attrition (Phillips & Fagan, 1982; Zamostny, Corrigan, &

Eggert, 1981). Whereas others have found satisfaction with

services to distinguish persisters from premature terminators

(Cochran & Stamler, 1989; Greenfield, 1983; Kokotovic &

Tracey, 1987; McNeill et al., 1987). Mennicke et al.

attribute this discrepancy to the use of different measures

used to assess client satisfaction. Most studies finding a

relationship between satisfaction and attrition used versions

of the Client Satisfaction Questionnaire (Larsen et al.,

1979), which is a well-validated measure of client

satisfaction with counseling services.


Satisfaction with Counseling


Tanner (1981) conducted a review of quantitative

research in which he studied the factors influencing client

satisfaction with mental health services. This review










summarizes the results of 38 studies which explored the

relationship of 31 variables to client satisfaction.

Variables included client and therapist demographic

variables, and interaction of these variables; behavioral,

stylistic or characterological variables, such as personality

and warmth; miscellaneous variables, such as therapist

assignment and duration of treatment; and finally, outcome or

effectiveness measures.

No client or therapist demographic variables were found

to consistently affect client satisfaction. Two client

characterological variables were found to be significantly

linked to satisfaction. Moberg (1977) found that alcoholic

inpatients were less satisfied on four of 12 scales when they

failed to identify themselves as alcoholics, and on six

scales when they planned to control their drinking rather

than abstain. However, these findings have not been

replicated.

With regard to therapist behaviors and stylistic

variables, clients appear to be more satisfied with

therapists that are active (Anderson, Harrow, Schwartz, &

Kupfer, 1972; Bent, Putnam, Kiesler, & Nowicki, 1976; Derita,

1976), warm (Bent, et al., 1976; McClanahan, 1974) empathic

(Anderson, et al., 1972; Fretz, 1966; McClanahan, 1974;

McNally, 1973) and interested (Anderson, et al., 1972; Fretz,

1966; McClanahan, 1974) in them. It is critical to note,

however, that these characteristics were measured by client

report.









Other significant therapist variables not obtained

through client report were social interest and profanity.

Clients' satisfaction increased as counselor social interest

rose (Zarski, Sweeney, & Barcikowski, 1977). Furthermore, an

experimental manipulation study found that middle class

clients were less satisfied with counselors who used

profanity (Heubush, & Horan, 1977). However, neither of

these studies have been replicated.

Two miscellaneous variables that were significantly

related to satisfaction were type of termination and duration

in treatment. Five studies examining termination type found

that outpatients who self-terminated were less satisfied with

counseling than those who remained in treatment (Balch,

Ireland, McWilliams, & Lewis, 1977; Denner & Halprin, 1974a,

1974b; Kline, Adrian, & Spevak, 1974; Larsen, 1978). In

addition, three studies of outpatients reported a positive

relationship between length of stay in treatment and at least

one satisfaction measure, such that the longer that patients

stayed in treatment, the more satisfied they were (Brown &

Manela, 1977; Frank, Salzman, & Fergus, 1977).

Finally five studies of clients' reports of treatment

effectiveness found a positive relationship to client

satisfaction (Edwards, Yarvis, Mueller, & Langsley, 1978;

Larsen, 1978; Larsen, et al., 1979; Simons, Wade, Morton, &

McSharry, 1978). However, there has been a lack of support

for these findings from related measures. Furthermore,

social desirability may account for a large part of both the







43

satisfaction and outcome ratings. Moreover, client report of

effectiveness of treatment typically accounts for less than

half the variance in satisfaction.

Freeman (1989) examined the influence of client and

counselor characteristics on satisfaction with counseling.

Subjects were 38 clients at a small liberal arts college

counseling center who had been seen for counseling. After

the termination interviews, clients and counselors completed

a questionnaire packet including the Client Satisfactions

Questionnaire (CSQ; Larsen et al., 1979) and two copies of

the Counselor Rating Form (CRF; Corrigan & Schmidt, 1983).

Participants were asked to rate first the other person and

then themselves on the CRF. Only 23 clients returned the

questionnaires.

Both counselors and clients perceived the counselor as

more attractive, expert, and trustworthy than the clients.

Clients rated themselves higher on the CRF dimensions than

the counselors rated their clients. Clients' attractiveness,

expertness and trustworthiness were significantly related to

the counselor's perception of clients' satisfaction. Client

satisfaction with services was significantly related to

client perception of counselor source characteristics and to

client perception of their own attractiveness.

It appears, therefore, that ratings of other was more

highly related to satisfaction than ratings of oneself. Only

in the case of clients' ratings of their own attractiveness

was self-rating significantly correlated with satisfaction.








44

However, in every instance, rating of other was significantly

and positively correlated with satisfaction.

Finally, a study by Neimeyer and Gonzales (1983)

investigated the duration, satisfaction, and perceived

effectiveness of cross-cultural counseling. Forty-nine white

and 21 nonwhite (8 Asian, 7 Black, 3 Hispanic, 1 Native

American, 2 unspecified) clients at a university counseling

center were seen by 13 white and 7 nonwhite (4 Black, 1

Asian, 2 Hispanic) counselors. Counselors and clients

completed a standard Counseling Services Evaluation Form

including demographic data, severity of the presenting

problem, duration of counseling and perceptions concerning

overall satisfaction and perceived effectiveness of services

at intake and 10 weeks after the initial session. Clients

were randomly assigned to therapists, however protesting was

performed to determine the equivalence of the four groups

(counselor race x client race) in terms of age and perceived

severity. No significant difference were found.

Results showed that White counselors provided fewer

sessions than nonwhite counselors. Nonwhite clients

expressed lower satisfaction than White clients, however,

this finding needs to be interpreted cautiously since mean

differences were minimal (M = 4.8, White; M = 4.3, nonwhite).

This result indicates that satisfaction did not vary as a

function of counselor race and was not related to cross-

racial differences. No difference in perceived effectiveness

of counseling was found among the four treatment dyads.







45

Neither counselors nor clients perceived any difference in

the effectiveness of counseling due to racial similarity or

dissimilarity. Therefore, it was concluded that race was not

a barrier to counseling.
















CHAPTER 3
METHODOLOGY


Subjects


Subjects were 75 students and two student spouses at the

University of Florida. The sample included clients being

seen individually (n = 64, 83%), in couples counseling (n =

6, 8%), and in group counseling (n = 7, 9%). Ages ranged

from 17 to 48, with an average age of 24.5 years. However,

62 percent of the sample (n = 47) were between the ages of 17

and 23. There were 58 females and 19 males in the sample.

The ethnic breakdown included 65 Caucasians (84%), four

African-Americans (5%), four Asians/Pacific Islanders (5%),

two Hispanics (3%), and two clients who did not report their

ethnic identity (3%). The sample consisted of graduate

students (n = 25, 32%), juniors (n = 19, 25%), sophomores (n

= 13, 17%), seniors (n = 12, 15%), freshmen (n = 4, 5%), post

baccalaureate students (n = 2, 3%), and student spouses (n =

2, 3%). Fifty-eight percent (n = 45) of the sample had been

in counseling previously, whereas, 42% (n = 32) were

attending counseling for the first time.

Twenty-six counselors (1 = 16 females, 10 males) from

the University of Florida Counseling Center, as well as,

three counselors (n = 3 males) from the University of Florida










Student Mental Health Services participated in the study.

The counselors from the Counseling Center consisted of seven

practicum students, nine interns, three counseling

associates, and seven senior staff members. The counselors

from Student Mental Health Services were a practicum student,

an intern, and a psychiatrist.


Instruments


The Expectations About Counseling-Brief Form (EAC-B;

Tinsley et al., 1980) consists of 66 items that are answered

on a 7-point Likert scale with response options which range

from not true to definitely true (Appendix A). Higher scores

indicate more positive expectations about counseling. Each

item is prefaced by the phrases "I expect to" or "I expect

the counselor to". For the purposes of this study only 53

items were used. The items excluded comprise the Realism

scale which is the most experimental scale of the inventory

and is not included in the factor scores. The remaining 17

scales of the EAC-B measure expectancies in four general

areas: Client Attitudes and Behaviors; Counselor Attitudes

and Behaviors; Counselor Characteristics; and Counseling

Process and Outcome. The internal consistency reliabilities

of the scales, based on the responses of 446 undergraduate

students ranged from .69 to .82 with a median reliability of

.76 (Tinsley, 1982). Test-retest reliabilities for a 2-month

interval ranged from .47 to .87 with a median of .71. All

but the responsibility scale had a test-retest reliability of










.60 or higher. The responsibility scale had a test-retest

reliability of .47 (Tinsley, 1982).

A factor analysis of the client expectancies measured by

the EAC-B (Tinsley et al., 1980) resulted in four factors:

Personal Commitment (EAC-B1), Facilitative Conditions (EAC-

B2), Counselor Expertise (EAC-B3), and Nurturance (EAC-B4).

These factors were used in the data analyses for this study.

The Personal Commitment factor reflect clients' expectancies

to be responsible, open, and motivated, to have an attractive

counselor, to have a counseling experience characterized by

concreteness and immediacy, and to experience a good outcome.

The Facilitative Conditions factor concerns expectancies that

the counselor will be genuine, trustworthy, accepting, and

tolerant, that the counselor will sometimes confront the

client, and also that the counseling experience will be

characterized by concreteness. The Counselor Expertise

factor reflects client's expectancies that the counselor will

by directive, empathic, and an expert. Lastly, the

Nurturance factor concerns client's expectancies that the

counselor will be accepting, self-disclosing, nurturant, and

attractive. The EAC-B also includes questions related to

demographic data, including: age, gender, race, year in

school, and whether the client has seen a professional

counselor before.

The Likert Trust Scale (LTS) was developed for use in

this study (Appendix B). It consists of four questions

answered on a 5-point Likert scale. Two questions relate to










the client's perception of his or her own trustworthiness

(Trustl), and ability to trust others (Trust2). The other two

questions relate to the client's perception of his or her

counselor's trustworthiness (Trust3), and ability to trust

others (Trust4). The response options range from 1-

trustworthy to 5-untrustworthy, and 5-trusting to 1-not

trusting. This type of instrument has been used in previous

research (Strong & Schmidt, 1970; Rothmeier & Dixon, 1980),

however, reliability data was not reported.

The Counseling Evaluation Inventory (CEI; Linden et al.,

1965) consists of 21 Likert items that assess client

satisfaction with counseling (Appendix C). Three scales were

identified through factor analysis : Counseling Climate

(e.g., "I distrusted the counselor"), Counselor Comfort

(e.g., In opening our conversations, the counselor was

relaxed and at ease"), and Client Satisfaction (e.g., I felt

satisfied as a result of my talks with the counselor").

Test-retest reliability estimates over a 14-day period ranged

from .63 to .78 for the three factors, and .83 for the total

inventory. Congruent validity has been demonstrated based on

a positive relationship (2 < .05) between practicum grades

for counselors and their clients' ratings on the CEI (Linden,

et al., 1965). Two scoring methods have been used with this

instrument a weighted scoring system and a standard 5-point

Likert scoring method (1 = always, 5 = never). The Likert

method of scoring, and the total inventory score were used in

this study.










The Client Satisfaction Ouestionnaire (CSQ; Larsen et

al., 1979) is an eight-item questionnaire used to assess

client satisfaction with mental health services (Appendix D).

Internal consistency for the CSQ is reported to be .87 for

university counseling center populations (Greenfield, 1983).

Scores range from 8 to 32, with higher scores indicating

higher satisfaction with services. Therapists' estimates of

how satisfied they believed the client to be were correlated

.56 (2 < .01) (Larsen et al., 1979) with the actual client

rating on the CSQ. This finding provides some evidence of

the scale's concurrent validity.

The Counselor Summary Sheet (CSS) was developed for use

in this study (Appendix E). This form was completed by the

counselors and was used to record the number of sessions

attended by clients (excluding intake), date of termination

(if applicable), reasons) for termination, and clients'

status: continuing counseling, mutual termination, dropout,

or other. Reasons for termination included: client

satisfactorily met his/her goals for counseling, counselor

did not have the skills/resources necessary to help the

client, client's life situation improved on its own, client

sought counseling elsewhere, counselor and client could not

agree on how to approach the client's problems, counselor and

client were not compatible, referred out because the client

needed/wanted more counseling than the Center could provide,

and other reason.










Mutual terminators were those individuals who planned a

termination with their counselors regardless of the number of

counseling sessions they had attended. Dropouts were those

individuals who terminated counseling without the knowledge

or against the advice of their counselor, and who felt that

their goals in counseling were not met.

The Client Summary Sheet (C1SS) was completed by the

clients (Appendix F). It is a parallel form of the Counselor

Summary Sheet. However, additionally, clients were asked to

indicate the degree to which they perceived themselves to be

similar to their counselors by circling a number on a likert

scale ranging from 1 not at all similar to 5 very

similar.


Procedure


Participants were 106 students or student spouses who

sought counseling at the University of Florida Counseling

Center (CC) and five who sought therapy at the University of

Florida Student Mental Health Services (SMHS). Of this

original sample, 77 (69%) subjects actually received

counseling beyond an initial, evaluation interview (intake)

(CC: 74, SMHS: 3). Of the other 34 students, 14 (41%) were

closed after their intake, 11 (32%) were referred to other

facilities on campus or to private practitioners in the

community, and nine (27%) were closed after being on the

waiting list. Walk-in clients, emergency clients, and

clients screening for a particular group did not take part in










the study. In addition, clients seeking career counseling

with the Peer Counselors were not asked to participate in

this study.

When clients arrived for an initial, evaluation

interview (intake) they were given a cover letter (Appendix

G) introducing them to the study, two copies of the Informed

Consent Form (Appendix H), the Expectations About Counseling-

Brief Form (EAC-B), the Likert Trust Scale (LTS), and an

envelope, along with the usual intake forms used by the

Counseling Center (CC) or Student Mental Health Services

(SMHS).

The Informed Consent Form included information about the

purpose of the study, the anticipated risks/benefits of

participation in the study, information about the principal

investigator, and information about the procedure to be used

in the study. Specifically, clients were informed that they

would be asked to complete two questionnaires prior to

beginning counseling, four questionnaires after their first

counseling session (including retaking the first two

completed), and five questionnaires after completing

counseling or after their sixth session (whichever was

first). The latter five questionnaires included retaking the

four completed previously. Completing all five

questionnaires was estimated to take 15 minutes.

Moreover, clients were informed that their participation

in the study was voluntary, and that their refusal to

participate in the study would not affect their access to










counseling. However, the Informed Consent Form also stated

that clients who completed the study would have their names

entered into a lottery from which ten names would be chosen.

The individuals selected would receive a money order for ten

dollars each.

Clients who agreed to participate in the study completed

the Expectations About Counseling-Brief Form and the Likert

Trust Scale prior to their intake and returned these

questionnaires to the receptionist sealed in the envelope

provided, along with one signed copy of the Informed Consent

Form. The other copy of the Consent Form was retained by

each client for his/her records. Those clients who decided

not to participate in the study simply returned the blank

questionnaires to the receptionist.

Upon receipt of the sealed envelope containing the first

two completed questionnaires and the Informed Consent Form,

the receptionist at the Counseling Center stapled a

memorandum to the inside of the client's folder containing

the usual intake forms used by the Counseling Center. After

the intake, the folder was either retained by the counselor

if the counselor had decided to pick up this particular

client for ongoing counseling, or the folder was placed in a

waiting list filing cabinet. The memorandum stapled to the

inside of the client's folder (Appendix I) was used by the

counselors picking up the client to inform the principal

investigator of the date of the client's first scheduled

counseling session, the name of the counselor, and the









client's address. The receptionist forwarded the envelopes

with the completed questionnaires, and consent forms to the

principal investigator. Clients were scheduled for a first

counseling session one or more weeks after their intake.

At Student Mental Health Services there is no waiting

list, such that each client who received an initial,

evaluation interview (intake) and was not referred out

immediately, was scheduled for a first session one week after

his/her intake. Therefore, on a specific day each week the

principal investigator collected the envelopes containing the

completed Expectations About Counseling-Brief Form (EAC-B),

Likert Trust Scale (LTS) questionnaires, and signed consent

forms completed prior to intake, and left envelopes

containing the second set of questionnaires for the

appropriate clients returning the following week for their

first scheduled counseling session.

The questionnaires completed prior to intake were

numerically coded upon receipt by the principal investigator

and separated from the signed Informed Consent Form. Clients

were asked not to write their names on the questionnaires.

Subsequent questionnaires were also numerically coded. This

procedure was used to ensure client confidentiality, and to

match questionnaires completed prior to intake with those

completed by the clients later in the study.

After each client's first counseling session, he/she was

either mailed or hand delivered (by the counselor), an

envelope containing a cover letter, (Appendix J) the










Expectations About Counseling-Brief Form (EAC-B) the Likert

Trust Scale (LTS),. the Client Satisfaction Questionnaire

(CSQ), and the Counseling Evaluation Inventory (CEI).

Clients were asked to complete and return these

questionnaires sealed in the envelope provided when the

client returned for his/her second counseling session.

Clients who did not return their questionnaires at the

appropriate time were mailed a reminder (Appendix K) and

subsequently mailed another set of questionnaires (n = 35,

54%).

The above procedure was also used for the third

administration which occurred after each client's sixth or

last session. Along with the four questionnaires completed

at the second administration, clients were also sent a cover

letter (Appendix L), and the Client Summary Sheet and asked

to complete and return these questionnaires to the principal

investigator. Clients who did not return their

questionnaires at the appropriate time were mailed a reminder

(n = 46, 60%). Clients who continually forgot to return

their questionnaires were telephoned by the principal

investigator and subsequently mailed another set of

questionnaires or came in to the Counseling Center or Student

Mental Health Services to complete questionnaires that were

left at the receptionist desk (n = 29, 38%). Also, after

their clients' sixth or last session, counselors were asked

to complete the Counselor Summary Sheet and return it to the

principal investigator.










After the termination of the project, the names of all

those clients who completed the project were entered into a

lottery. Ten names were chosen at random and ten dollar

money orders were mailed to these individuals. Data was

collected over a period of two semesters, which was

approximately 28 weeks.


Hypotheses


This study examined clients' expectations of their

counselors and the counseling process, and clients' levels of

perceived trust and trustworthiness prior to beginning

counseling. Moreover, changes in these variables and

clients' satisfaction with counseling were investigated in

relation to clients' actual experiences of counseling.

Specific hypotheses were:

1) Clients will exhibit higher positive expectations

(Personal Commitment, Facilitative Conditions, Counselor

Expertise, Nurturance) of their counselors and counseling

after their first session than prior to beginning counseling.

2) Clients will evidence higher levels of perceived

trustworthiness and trust (Client Trustworthiness, Client

Trust, Counselor Trustworthiness, Counselor Trust) after

their first session than prior to beginning counseling.

3) Clients will exhibit higher positive expectations of

their counselors and counseling after two or more sessions

than prior to beginning counseling or after their first

session. There will be a significant positive association







57

between number of counseling sessions attended by clients and

clients' levels of expectations, such that as the number of

sessions increases so will clients' expectations for Personal

Commitment, Facilitative Conditions, Counselor Expertise, and

Nurturance.

4) Clients will evidence higher levels of perceived

trustworthiness and trust after two or more sessions than

prior to beginning counseling or after their first session.

There will be a significant positive association between the

number of counseling sessions attended by clients, and both

clients' reported levels of their own (Client

Trustworthiness, Client Trust) and their counselors'

trustworthiness and trust (Counselor Trustworthiness,

Counselor Trust).

5) As clients' expectations for counseling increase,

clients' satisfaction with counseling (Client Satisfaction;

SATIS) and with their counselors (Counseling Evaluation;

CEVAL) will also increase.

6) As clients' perceived levels of their own and their

counselors' trustworthiness and trust increase, clients'

satisfaction with counseling (SATIS) and with their

counselors (CEVAL) will also increase.

7) Compared to clients who plan a termination with

their counselors, clients who drop out of counseling will

evidence lower expectations of counseling (Personal

Commitment, EAC-Bl; Facilitative Conditions, EAC-B2;

Counselor Expertise, EAC-B3; Nurturance, EAC-B4) and lower









levels of perceived trustworthiness and trust (Client

Trustworthiness, Client Trust, Counselor Trustworthiness,

Counselor Trust) prior to beginning counseling. In addition,

dropouts will have less satisfaction with counseling (SATIS)

and with their counselors (CEVAL) than mutual terminators.

8) Compared to clients' expectations (EAC-B) and levels

of perceived trust/trustworthiness (TRUST) prior to beginning

counseling (timel), clients' expectations, levels of

trust/trustworthiness, and satisfaction with counseling

(SATIS) and with their counselors (CEVAL) early in counseling

timee) will be better predictors of client satisfaction with

counseling/counselors later in counseling timee3.

9) There will be significant positive associations

between clients' levels of perceived similarity to their

counselors (SIM) and clients' satisfaction with their

counseling (Client Satisfaction; SATIS) and their counselors

(Counseling Evaluation; CEVAL).















CHAPTER 4
RESULTS


The major dependent variables of study in this research

were the following: clients' satisfaction with counseling

(SATIS) and clients' satisfaction with their counselors

(CEVAL). The major independent variables were: clients'

expectations about counseling (EAC-B), clients' perceptions

of their own and their counselors' trust/trustworthiness

(Trust), and time in counseling. Clients' expectations were

comprised of four factors: Personal Commitment (EAC-B1),

Facilitative Conditions (EAC-B2), Counselor Expertise (EAC-

B3), and Nurturance (EAC-B4). Clients' perceptions of their

own and their counselors' trust/trustworthiness included four

variables: Client Trustworthiness (Trustl), Client Trust

(Trust2), Counselor Trustworthiness (Trust3), and Counselor

Trust (Trust4). Time in counseling was divided into three

periods: precounseling (timel), after the first counseling

session timee) and after the sixth or last counseling

session timee3.

Due to the inability to control when clients' actually

returned their questionnaires, not all questionnaires were

returned after the first and prior to the second counseling

session for the second administration of the questionnaires,

nor after the sixth and prior to the seventh counseling









session for the third administration of the questionnaires.

Therefore, the first administration of the questionnaires

(timel) was prior to beginning counseling, and thus the

number of sessions attended was zero. For the second

administration timee) the number of sessions attended ranged

from one to five sessions (m = 1.11, SD = 1.38). For the

third administration timee) the number of sessions attended

ranged from one to 19 sessions (m = 4.50, SD = 4.37). Four

clients dropped out of counseling after their first session

and therefore were administered the third set of

questionnaires rather than the second set of questionnaires.

The data for these four clients were not included in the

analyses for the first four hypotheses.

In order to maintain the integrity of the study, only

data obtained from those clients returning their second set

of questionnaires after their first counseling session and

prior to their second session were included in the analyses

for the first four hypotheses (q = 24, 53%). The analyses

for these hypotheses were repeated measures analyses of

variance (ANOVAs) and repeated measures analyses of

covariance (ANCOVAs). Repeated measures designs employ the

use of a subject as his/her own control and follow-up.

Therefore it was necessary to use only the appropriate data

in order to obtain the comparisons stated in the first four

hypotheses: prior to counseling, after the first session,

and after two or more sessions.










To test hypotheses five through nine, regressions,

correlations, and analyses of variance (ANOVAs) procedures

were used. All data obtained regardless of when the

questionnaires were returned could be and were used in these

analyses. The return rates were as follows: time = 73%, n

= 45; time = 83%, n = 64.


Preliminary Analyses


It has been noted that clients come in to counseling

with preformed ideas about their counselor and the counseling

process (Gelso & Carter, 1985). It appeared possible,

therefore, that clients' previous counseling experiences

(PREV) may affect their expectations and perceptions of their

present counselors and counseling experiences. It is further

noted that there have been inconsistent findings in the

counseling research literature pertaining to the influence of

actual counselor experience level (CEXP) on the counseling

process (Heppner & Heesacker, 1982; Auerbach & Johnson,

1977).

Thus, prior to performing the statistical analyses to

test the stated hypotheses, a series of repeated measures

analyses of variance (ANOVAs) were performed to determine

whether the dependent measures covary with clients' previous

counseling experience (PREV: 1 = yes; 2 = no), counselors'

experience levels (CEXP: 1 = prepracticum; 2 = practicum; 3

= intern/associate; 4 = senior staff; 5 = psychiatrist), or

the interaction of these two variables (PREV*CEXP). The









results indicated whether these variables (PREV, CEXP,

PREV*CEXP) needed to be included as covariates in the

analyses to test the stated hypotheses.

Specifically, separate repeated measures ANOVAs were

performed with each of the following variables as the

dependent measure: the four Expectations About Counseling

(EAC-B) factors [Personal Commitment (EAC-B1), Facilitative

Conditions (EAC-B2), Counselor Expertise (EAC-B3), and

Nurturance (EAC-B4)] at each of the three time periods, the

four Likert Trust Scale variables [Client Trustworthiness

(Trustl), Client Trust (Trust2), Counselor Trustworthiness

(Trust3), Counselor Trust (Trust4)] at each of the three time

periods, and the two satisfaction measures [Client

Satisfaction (SATIS), Counseling Evaluation, (CEVAL)] at each

of the last two time periods. The independent variables in

each of these repeated measures ANOVAs were time in

counseling (timel, time time3, clients' previous counseling

experience (PREV), counselors' experience levels (CEXP), and

the interaction of these latter two variables (PREV*CEXP).

In the following sections are the results of the two repeated

measures ANOVAs that proved significant the one with

Counselor Trustworthiness (Trust3) as the dependent variable,

and the one with Counseling Evaluation (CEVAL) as the

dependent variable.










The Repeated Measures Analysis of Variance with Counselor
Trustworthiness as the Dependent Variable


A repeated measures analysis of variance (ANOVA) was

performed with Counselor Trustworthiness (Trust3) as the

dependent measure,and the following as independent variables:

counseling received over three time periods (timel, time,

time3, clients' previous counseling experience (PREV),

counselor experience level (CEXP) and the interaction of the

latter two variables (PREV*CEXP). Results of the between

subjects test revealed a significant overall effect for

PREV*CEXP, F(l, 38) = p < .031. This interaction (PREV*CEXP)

indicates that clients' perceptions of their counselors as

trustworthy were affected by the counselors' experience

levels and whether the client had previously attended

counseling. Those clients with previous counseling

experience did not perceive their counselors as significantly

more or less trustworthy based on the counselors' experience

levels. Whereas, those clients with no previous counseling

experience perceived the more experienced counselors as

significantly less trustworthy than the less experienced

counselors.

Results of the within subjects test revealed a

significant time*CEXP interaction, F(2, 76) = 3.95, p < .023,

indicating that the relationship between Counselor

Trustworthiness (Trust3) and counselors' experience levels

(CEXP) was not the same across the three time periods. There










was no relationship between Counselor Trustworthiness

(Trust3) and counselors' experience levels (CEXP) prior to

beginning counseling (timel). However, early in counseling

timeme, m = 1.11 sessions), F(3, 38) = 3.05, p :.0401], and

later in counseling timeme, m = 4.5 sessions), (3, 38) =

4.88, p S .0058], the more experienced counselors were seen

as less trustworthy.

Pearson correlations confirmed that there was a moderate

but significant negative relationship between CEXP and Trust3

at time (r = -.34, p < .02). The correlations between CEXP

and Trust3 at timel (E = -.04, p < .67) and time (1 = -.18,

p < .16) were also negative, however neither were

significant. These results indicate that as the counselors'

experience levels increased, clients' perceptions of their

counselors as trustworthy decreased. Plots confirmed these

finding showing negative slopes between CEXP and Trust3 at

time, and at time.

Interestingly, the interaction between clients'

previous counseling experience and counselors' experience

levels (PREV*CEXP) was not detected in the within subjects

analysis, although the trend did increase through the three

time periods. Since the effect was found in the between

subjects analysis, the current study may not have had enough

statistical power to detect the three-way interaction

(time*PREV*CEXP).











The Repeated Measures Analysis of Variance with Counseling
Evaluation as the Dependent Variable


A repeated measure ANOVA was performed with Counseling

Evaluation (a measure of clients' satisfaction with their

counselors; CEVAL) as the dependent measure, and the

following as independent variables: counseling received over

two time periods timee, time3, clients' previous counseling

experience (PREV), counselors' experience levels (CEXP), and

the interaction of these two variables (PREV*CEXP). The

between subjects test revealed a significant overall effect

for CEXP, F(l, 37) = 3.86, p < .056. Apparently the

counselors' experience levels also affected the clients'

satisfaction with their counselors.

The within subjects test revealed a significant

time*CEXP*PREV interaction, F(l, 37) = 4.26, a < .0462.

There was no significant effect for CEVAL at time. However,

there was a significant effect for CEVAL at time, (3, 37) =

3.99, Ep .0147. Moreover, CEXP*PREV (p < .0270) was

significantly associated with Counseling Evaluation (CEVAL)

at time. These results indicate that counselors' experience

levels (CEXP) and clients' previous counseling experience

(PREV) were not significant factors in clients' satisfaction

with their counselors (CEVAL) early in counseling timee2.

However, later in counseling timee) an interaction between

CEXP and PREV apparently became impactful in clients'

satisfaction with their counselors (CEVAL).







66

Pearson correlations and plots were conducted separately

for clients with no previous counseling experience and

clients with previous counseling experience. Pearson

correlations revealed that for clients with no previous

counseling experience, there were moderate, but

nonsignificant negative relationships between CEVAL and CEXP

both at time (I = -.42, p s .108), and time (r = -.33, 2

.103). However, for clients with previous counseling

experience, the relationship between CEXP and CEVAL changed

from being negative at time (r = -.10, p < .627) to being

positive at time (r = .28, E < .094). These correlations

were also not significant. Plots confirmed these findings

showing consistent negative slopes at time and time for

clients with no previous counseling experience. In contrast,

for clients with previous counseling experience, plots showed

a negative slope between CEVAL at time and CEXP, but a

positive slope between CEVAL at time and CEXP, which

resulted in the interaction effect.


Tests of the Hypotheses


Since the preliminary analyses revealed that both

Counselor Trustworthiness (Trust3) and Counseling Evaluation

(CEVAL) were significantly affected by clients' previous

counseling experience (PREy), and counselors' experience

levels (CEXP), these two variables and their interaction

(PREV, CEXP, and PREV*CEXP) were included as covariates in

the analyses involving Trust3 and CEVAL. Where results with









the covariates were very similar to results without the

covariates, the simpler models (without the covariates) are

reported.

As previously stated, hypothesis one was that clients

will exhibit higher positive expectations (expectation

factors: Personal Commitment, Facilitative Conditions,

Counselor Expertise, Nurturance) of their counselors and

counseling after their first session than prior to beginning

counseling. The analyses for this hypothesis were t-tests

for dependent samples. Results indicated no significant

differences between the means for the expectation factors

from prior to beginning counseling to after the first

counseling session, failing to support hypothesis one.

Hypothesis two stated that clients will also evidence

higher levels of perceived trust and trustworthiness (Client

Trustworthiness, Client Trust, Counselor Trustworthiness,

Counselor Trust) after their first session than prior to

beginning counseling. T-tests for dependent samples were

conducted for Client Trustworthiness (Trustl), Client Trust

(Trust2), and Counselor Trust (Trust4). Results revealed a

significant (somewhat borderline) difference for Trustl from

prior to beginning counseling to after the first counseling

session (t = -2.01, p < .0558). Means indicated that

clients' perceptions of themselves as trustworthy decreased

from prior to beginning counseling (m = 4.63, SD = 0.58) to

after their first session (m = 4.38, SD = 0.82). However,







68

there were no significant differences in the means for Trust2

or Trust4 over the same time period.

A repeated measures analysis of covariance (ANCOVA) was

conducted with Counselor Trustworthiness (Trust3) as the

dependent variable, time in counseling (timel, time) as the

independent variable, and clients' previous counseling

experience (PREV), counselors' experience levels (CEXP), and

the interaction of these two latter variables (PREV*CEXP) as

covariates. The within subjects test for the entire sample

indicated no overall time effect, and no differences in the

means for Counselor Trustworthiness (Trust3) over time due to

previous counseling experience (PREV), counselor experience

level (CEXP) or the interaction of these two variables

(PREV*CEXP). There was a decrease in clients' perceptions of

their counselors as trustworthy from prior to beginning

counseling (m = 4.71, SD = 0.46) to after their first session

(m = 4.21, SD = 0.98), however, the difference was not

significant. The results obtained failed to support

hypothesis two.

Hypothesis three stated that clients will exhibit higher

positive expectations for Personal Commitment (EAC-Bl),

Facilitative Conditions (EAC-B2), Counselor Expertise (EAC-

B3), and Nurturance (EAC-B4) after two or more sessions than

prior to beginning counseling or after their first session.

Moreover, there will be a significant positive association

between number of counseling sessions attended by clients and

clients' expectations about counseling, such that as number










of sessions increases so will clients' expectations for

Personal Commitment, Facilitative Conditions, Counselor

Expertise, and Nurturance (expectation factors).

To test hypothesis three, separate repeated measures

analyses of variance (ANOVAs) were conducted with each

expectation factor as a dependent measure, and counseling

received over three time periods as the independent measure.

Results of the four repeated measures ANOVAs indicated that

there were no overall effects due to time for any of the

expectation factors, indicating that there were no

significant differences in the means for any of the four

expectation factors from prior to beginning counseling, to

after the first or after two or more counseling sessions.

There was an increase in the means for Nurturance (EAC-B4)

over the three time periods (see Table 4-1, however, the

differences in the means were not significant.

In addition, Pearson correlations were utilized to

determine the type of relationship (positive vs negative)

between number of sessions attended by clients and clients'

expectation for Personal Commitment (EAC-B1), Facilitative

Conditions (EAC-B2), Counselor Expertise (EAC-B3), and

Nurturance (EAC-B4). Results indicated that none of the

correlations were significant, such that clients'

expectations about counseling were not significantly related

to the amount of time clients spent in counseling. The

results of the analyses conducted for hypothesis three,

therefore, failed to support the hypothesis.










Table 4-1

Simple Statistics for the Exoectation Factors


Variable Timela Time2b Time3c

n m SD n m SD n m SD


EAC-Bl 21 5.68 0.75 23 5.63 0.56 24 5.78 0.66

EAC-B2 22 5.41 0.92 24 5.30 0.99 24 5.38 0.76

EAC-B3 24 4.01 1.13 24 3.86 1.43 24 4.19 1.34

EAC-B4 22 4.53 1.03 24 4.59 1.19 24 4.62 1.05



a0 sessions, bl session, cm = 6.33 sessions, SD = 3.45

EAC-B: 7 point Likert scale


Hypothesis four was that clients will also evidence

higher levels of perceived trustworthiness and trust [Client

Trustworthiness, (Trustl); Client Trust, (Trust2); Counselor

Trustworthiness, (Trust3); Counselor Trust, (Trust4)] after

two or more sessions than prior to beginning counseling or

after their first session. Moreover, there will be a

significant positive association between number of counseling

sessions attended by clients and clients' reported levels of

their own trustworthiness and trust as well as clients'

levels of perceived counselor trustworthiness and trust.

The analyses for hypothesis four involved separate

repeated measures ANOVAs with Trustl, Trust2, and Trust4 as

dependent variables, and a repeated measures ANCOVA with

Trust3 as the dependent variable. Counseling received over









three time periods was the independent variable. In

addition, Pearson correlations were utilized to determine the

type of relationship (positive vs negative) between number of

sessions attended by clients and clients' reports of their

own trustworthiness and trust, as well as clients' perceived

counselor trustworthiness and trust. Results of the three

repeated measures ANOVAs with Trustl, Trust2 and Trust4 as

the dependent measures revealed that there were no overall

time effects for any of the three trust variables, indicating

that there were no significant differences in the means for

any of the three Trust variables from prior to beginning

counseling, to after the first or after two or more

counseling sessions. There was an increase in the means for

Counselor Trust (Trust4) over the three time periods (Table

4-2). However, the differences in the means were not

significant.

A repeated measures ANCOVA was conducted with Trust3

(Counselor Trustworthiness) as the dependent measure,

counseling received over three time periods (timel, time,

time) as the independent measure, and clients' previous

counseling experience (PREV), counselors' experience levels

(CEXP), and the interaction of these two variables

(PREV*CEXP) as covariates. The within subjects test of the

repeated measures ANCOVA with Counselor Trustworthiness as

the dependent variable for the entire sample revealed that

there was no overall time effect, nor significant differences

in the means for Trust3 over time due to clients' previous










counseling experience (PREV), the counselors' experience

levels (CEXP) or the interaction of these two variables

(PREV*CEXP).


Table 4-2

Simple Statistics for the Trust Variables



Variable Timela Time2b Time3c

n m SD n m SD n m SD


Trustl 24 4.63 0.58 24 4.38 0.82 24 4.54 0.59

Trust2 24 3.58 0.78 24 3.33 0.92 24 3.33 0.96

Trust3 24 4.71 0.46 24 4.21 0.98 24 4.63 0.71

Trust4 24 3.75 0.74 23 3.87 0.97 24 4.17 0.76



a0 sessions, bl session, cm = 6.33 sessions, SD = 3.45

Trust: 5 point Likert scale


Results of the Pearson correlations between number of

sessions attended by clients and the Trust variables revealed

that none of the correlations were significant, such that

clients' perceptions of themselves and their counselors as

trustworthy and trusting were not significantly related to

the number of counseling sessions attended by the clients.

Therefore, the results of the analyses conducted for

hypothesis four failed to support the hypothesis.

Hypothesis five was that as clients' expectations for

counseling (expectation factors: Personal Commitment,







73

EAC-Bl; Facilitative Conditions, EAC-B2; Counselor Expertise,

EAC-B3; Nurturance, EAC-B4) increase, clients' satisfaction

with counseling (Client Satisfaction, SATIS) and with their

counselors (Counseling Evaluation, CEVAL) will also increase.

Both Pearson correlations between the expectation factors and

SATIS and CEVAL, and multiple regressions were conducted to

test this hypothesis. Only Personal Commitment (EAC-B1) at

time (m = 1.11 sessions) showed moderate but significant

positive correlations with clients' general satisfaction with

counseling (Client Satisfaction; SATIS), and with clients'

satisfaction with their counselors (Counseling Evaluation;

CEVAL) at time. These results indicated that early in

counseling timee2, as clients' expectations for Personal

Commitment to counseling increased, so did their satisfaction

with their counseling (SATIS) and with their counselors

(CEVAL). See Table 4-3 for the results of the Pearson

Correlations.

Four multiple regressions were also conducted. More

specifically, the first regression included Client

Satisfaction (SATIS) at time (m = 1.11 sessions) as the

criterion variable and the following as predictor variables:

the four expectation factors (Personal Commitment, EAC-B1;

Facilitative Conditions, EAC-B2; Counselor Expertise, EAC-B3,

Nurturance, EAC-B4) at time. Results of this regression

revealed that the model itself was not statistically









Table 4-3

Pearson Correlation Coefficients Amona the Expectation
Factors and the Counselina/Counselor Satisfaction Measures



Variables Time2a Time3b



Client Satisfaction and

Personal Commitment .34* .14

Facilitative Conditions .17 -.10

Counselor Expertise .05 -.06

Nurturance .25 -.03

Counseling Evaluation and

Personal Commitment .43** .19

Facilitative Conditions .14 -.04

Counselor Expertise -.04 -.09

Nurturance .12 -.04


am = 1.11 sessions, SD = 1.38

*< p .05 **g < .005


bm = 4.5 sessions, SD = 4.37.


significant (E(4, 38) = 1.47, 2 s .2313) indicating that none

of the expectation factors at time were related to Client

Satisfaction (SATIS) at time. Likewise, the R-square (a2 =

.134) was quite small indicating that the expectation factors

at time did not account for much of the variance in SATIS at









time and, therefore, were poor predictors of Client

Satisfaction (SATIS) with counseling in general at time.

The second regression involved SATIS at time as the

criterion variable, and the four expectation factors at time

(m = 4.5 sessions) as the predictor variables. Results

revealed that the model itself was not significant (E(4, 59)

= 1.90, B .1232), suggesting that none of the expectation

factors at time were related to clients' satisfaction with

counseling (SATIS) at time. Moreover, the R-square (R2 =

.114) was quite small, and Pearson correlations confirmed

that there were minimal relationships between clients'

satisfaction with counseling (SATIS) at time and clients'

expectations about counseling at time (see Table 4-3)

indicating that the expectation factors at time accounted

for only a very small amount of the variance in SATIS at

time and therefore were not good predictors of this

variable.

The third regression included Counseling Evaluation

(CEVAL) at time as the criterion variable and the following

as predictor variables: the four expectations factors

(Personal Commitment, EAC-B1; Facilitative Conditions, EAC-

B2; Counselor Expertise, EAC-B3, Nurturance, EAC-B4) at time

(m = 1.11 sessions), clients' previous counseling experience

(PREV), counselors' experience levels (CEXP), and the

interaction of these two latter variables (PREV*CEXP).

Results of this regression (as shown in Table 4-4) indicated

that the model itself was significant (E(7, 35) = 2.39,










Table 4-4

Multiple Regression for Counseling Evaluation at Time2 with
the Expectation Factors at Time2 as Predictors



Source DF SS MS F


Model

Error

Total


EAC-B1-2

EAC-B2-2

EAC-B3-2

EAC-B4-2

PREV

CEXP

PREV*CEXP

Variable


Intercept

EAC-B1-2

EAC-B2-2

EAC-B3-2

EAC-B4-2

PREV

CEXP

PREV*CEXP


2.58

5.39

7.97

1.77

0.15

0.00

0.02

0.18

0.56

0.15

Parameter
Estimate

2.81

0.59

-0.13

-0.00

-0.05

-0.60

-0.26

0.18


0.37

0.15



1.77

0.15

0.00

0.02

0.18

0.56

0.15

Standard
Error

0.77

0.17

0.13

0.07

0.12

0.55

0.13

0.18


2.39






11.46

0.97

0.01

0.16

1.18

3.66

0.94

T


3.65

3.39

-0.99

-0.08

-0.40

-1.09

-1.98

0.97


B2 = .324 CV = 8.83 Root MSE = .392


.0414





.0018

.3303

.9389

.6920

.2838

.0639

.3386




.0008

.0018

.3303

.9389

.6920

.2838

.0559

.3386

= 4.45










p .0414), suggesting that at least one of the independent

variables was significantly related to clients' satisfaction

with their counselors (CEVAL) at time. Personal Commitment

(EAC-B1) was the only significant predictor variable (p S

.0018) in this model. Interestingly, the R-square obtained

was small (R2 = .324) indicating that the model accounted for

only a small amount of the variance in Counseling Evaluation

(CEVAL). However, Pearson correlations indicated that

Personal Commitment was also the only expectation factor

significantly correlated with CEVAL at time (r = .43, p

.004) suggesting that Personal Commitment may still be a good

predictor of Counseling Evaluation at time, if it (EAC-B1)

accounts for most or all of the variation explained by the

model. The fourth regression included CEVAL at time as the

criterion variable and the following as predictor variables:

the four expectation factors (EAC-B) at time (m = 4.5

sessions), clients' previous counseling experience (PREV),

counselors' experience levels (CEXP), and the interaction of

these two latter variables (PREV*CEXP). Results indicated

that the model was significant (E(7, 56) = 2.14, p S .0542),

suggesting that at least one of the independent variables was

significantly related to clients' satisfaction with their

counselors at time (see Table 4-5).

Personal Commitment (EAC-B1, p < .0081), clients'

previous counseling experience (PREV, p < .0205), and the

interaction of clients' previous counseling experience and











Table 4-5

Multiple Regression for Counseling Evaluation at Time3 with
the Expectation Factors at Time3 as Predictors



Source DF SS MS E


Model 7 5.59 0.80 2.14 .0542


Error

Total


20.94

26.53


EAC-B1-3

EAC-B2-3

EAC-B3-3

EAC-B4-3

PREV

CEXP

PREV*CEXP

Variable


Intercept

EAC-B1-3

EAC-B2-3

EAC-B3-3

EAC-B4-3

PREV

CEXP

PREV*CEXP


2.82

0.12

0.00

0.15

2.13

0.11

2.64

Parameter
Estimate

3.70

0.46

-0.13

-0.01

-0.11

-1.54

-0.34

0.56


0.37


2.82

0.12

0.00

0.15

2.13

0.11

2.64

Standard
Error

0.76

0.17

0.23

0.08

0.17

0.64

0.15

0.21


7.54

0.32

0.01

0.40

5.68

0.30

7.07

T


4.83

2.75

-0.57

-0.07

-0.63

-2.38

-2.21

2.66


E2 = .211 CV = 14.24 Root MSE = .611


.0081

.5709

.9418

.5287

.0205

.5875

.0102

E


.0001

.0081

.5709

.9418

.5287

.0205

.0312

.0102

m= 4.29










counselors' experience levels (PREV*CEXP, P < .0102) appeared

as significant predictors of Counseling Evaluation (CEVAL) at

time. To clarify the interaction effect (PREV*CEXP), the

parameter estimates revealed that both PREV (-1.54) and CEXP

(-0.34) were negatively associated with CEVAL. These

negative associations indicate that clients with previous

counseling experience were more likely to have higher

satisfaction with their counselors (CEVAL). In addition, it

appears that clients were more satisfied with counselors who

had less experience. The interaction between clients'

previous counseling experience and counselors' experience

levels (PREV*CEXP = 0.56) indicates that clients with

previous counseling experience had higher satisfaction with

more experienced counselors. In contrast, clients with no

previous counseling experience had higher satisfaction with

less experienced counselors. Once again, the R-square

obtained was quite small (R2 = .211), indicating that the

predictors in this model only accounted for a small amount of

the variance in CEVAL at time. Therefore, it is doubtful

whether even the significant predictors are very influential

in predicting Counseling Evaluation (CEVAL) at time.

Results of the multiple regressions for hypothesis five

revealed that only small amounts (R2 = .114 to .324) of the

variance in SATIS and CEVAL at time and time were accounted

for by the four expectation factors. These results indicate

that the expectation factors were not highly influential

predictors of clients' satisfaction with their counseling










(SATIS) and with their counselors (CEVAL). However, the

Pearson correlations did reveal significant positive

correlations between Personal Commitment (EAC-B1) at time

and both SATIS (r = .343, p < .0226) and CEVAL (r = .429, 2

.0040) at time, indicating that as clients' expectations for

their Personal Commitment to counseling increased, so did

clients' satisfaction with counseling and with their

counselors giving marginal support to hypothesis five for

EAC-B1

Hypothesis six was that as clients' perceived levels of

their own and their counselors' trustworthiness and trust

(Client Trustworthiness, Trustl; Client Trust, Trust2;

Counselor Trustworthiness, Trust3; Counselor Trust, Trust4)

increase, clients' satisfaction with counseling (Client

Satisfaction, SATIS) and with their counselors (Counseling

Evaluation, CEVAL) will also increase. Both Pearson

correlations between the trust variables and SATIS and CEVAL,

and multiple regressions were utilized to test the

hypothesis.

The Trust variables all showed significant positive

correlations with SATIS and/or CEVAL at one or both time

periods (see Table 4-6), indicating that both clients'

perceptions of themselves and perceptions of their counselors

as trustworthy and trusting were important in clients'

satisfaction with counseling (SATIS) and/or with their

counselors (CEVAL). Moreover, Counselor Trustworthiness










Table 4-6

Pearson Correlation Coefficients Among the Trust Variables
and the Counseling/Counselor Satisfaction Measures



Variables Time2a Time3b


Client Satisfaction and

Client Trustworthiness .37** .10

Client Trust .34* .15

Counselor Trustworthiness .60*** .60***

Counselor Trust .20 .43***

Counseling Evaluation and

Client Trustworthiness .28 .12

Client Trust .31* .10

Counselor Trustworthiness .67*** .64***

Counselor Trust .23 .43***


am = 1.11 sessions, SD = 1.38 m = 4.5 sessions, SD = 4.37.

*g < .05 **D < .01 ***P < .0005


(Trust3) evidenced the highest correlations with both SATIS

(r = .600, p < .0001) and CEVAL (r = .665, ] < .0001) at

time, and time (SATIS: r = .603, Ep .0001; CEVAL: r =

.636, p j .0001). These findings indicate that as clients'

perceptions of their counselors as trustworthy increased, so

did clients' satisfaction with counseling in general and

satisfaction with their counselors.










Four multiple regression were also performed. The first

multiple regression involved Client Satisfaction (SATIS) at

time as the criterion variable and the four Trust variables

at time [Client Trustworthiness, (Trustl); Client Trust,

(Trust2); Counselor Trustworthiness, (Trust3); Counselor

Trust, (Trust4)] as the predictor variables. Results

revealed that the model was significant (E(4, 38) = 6.91, 2 S

.0003), indicating that at least one of the Trust variables

was significantly related to clients' satisfaction with

counseling (SATIS) at time. Counselor Trustworthiness

(Trust3) was the only predictor that reached statistical

significance (p : .0013). The R-square for this model (R2 =

0.421) was moderate, indicating that the predictors in this

model accounted for some of the variance in SATIS at time,

however, over fifty percent of the variance was still

unaccounted for, indicating that collectively the Trust

variables were not very good predictors of SATIS at time.

However, considering the highly significant correlation

between Counselor Trustworthiness (Trust3) and SATIS at time

(r = .600, p c .0001), Trust3 may explain most of the

variation accounted for by the model, and thus may still be a

good predictor of SATIS at time. See Table 4-7 for the

results of this regression.

The second regression included Client Satisfaction

(SATIS) at time (m = 4.5 sessions) as the criterion

variable, and the four Trust variables at time as the











Table 4-7

Multiple Regression for Client Satisfacti
Trust Variables at Time2 as Predictors


Source DF SS MS E P


Model 4 207.88 51.97 6.91 .0003

Error 38 285.75 7.52

Total 42 493.63


Variable DF Parameter Standard I P
Estimate Error


Intercept 1 12.43 3.45 3.60 .0009

Trustl-2 1 0.51 0.70 0.72 .4751

Trust2-2 1 0.75 0.44 1.69 .0985

Trust3-2 1 2.14 0.62 3.46 .0013

Trust4-2 1 -0.16 0.54 -0.29 .7757

R2 = .421 CV = 10.58 Root MSE = 2.74 m = 25.91

Adj. R2 = .360




predictor variables (see Table 4-8). Again results revealed

that the model itself was significant (E(4, 59) = 8.58, s

.0001), and Counselor Trustworthiness (Trust3) was the only

significant predictor (p < .0001). However the R-square for

this model (R2 = .368) was small. Once again it appears that

collectively the predictors in this model only accounted for


on at T me2 w th the


i i










Table 4-8

Multiple Regression for Client Satisfaction at Time3 with the
Trust Variables at Time3 as Predictors



Source DF SS MS F P



Model 4 706.06 190.02 8.58 .0001

Error 59 1306.92 22.15

Total 63 2066.98


Variable DF Parameter Standard T P
Estimate Error


Intercept 1 4.65 5.49 0.85 .4004

Trustl-3 1 -0.17 1.05 -0.16 .8705

Trust2-3 1 0.05 0.58 0.09 .9296

Trust3-3 1 4.21 1.03 4.09 .0001

Trust4-3 1 0.48 0.85 0.56 .5748

R2 = .368 CV = 18.81 Root MSE = 4.71 m = 25.91

Adj. R2 = .325




a small amount of the variance in SATIS at time. Yet, it is

possible that because Trust3 was highly correlated with SATIS

at time (r = .603, p S .0001), Trust3 accounted for most of

the variance in the model and may still be influential in

predicting SATIS at time.

The third regression involved Counseling Evaluation

(CEVAL) at time (m = 1.11 sessions) as the criterion










variable, and the four Trust variables at time [Client

Trustworthiness, (Trustl); Client Trust, (Trust2); Counselor

Trustworthiness, (Trust3); Counselor Trust, (Trust4)] as the

predictor variables (see Table 4-9). Once more, this model


Table 4-9

Multiple ReGression
the Trust Variables


for Counseling Evaluation at Time2 with
at Time2 as Predictors


Source DF SS MS F 2



Model 4 3.55 0.89 8.10 .0001

Error 38 4.16 0.11

Total 42 7.72


Variable DF Parameter Standard T p
Estimate Error

Intercept 1 2.95 0.42 7.07 .0001

Trustl-2 1 -0.03 0.08 -0.38 .7073

Trust2-2 1 0.06 0.05 1.21 .2346

Trust3-2 1 0.34 0.07 4.51 .0001

Trust4-2 1 -0.02 0.06 -0.24 .8074

R2 = .460 CV = 7.47 Root MSE = .331 m = 4.43

Adj. R2 = .403




was also statistically significant ((4, 38) = 8.10, p < =

.0001), and consistently, Counselor Trustworthiness (Trust3)

was the only predictor variable to reach statistical


the Trust Variables








86

significance (2 s .0001). Results revealed that the R-square

for this model (R2 = .460) was moderate, indicating that 46%

of the variance in CEVAL at time was accounted for by the

Trust variables. However, compared to the other trust

variables, Counselor Trustworthiness (CEVAL) appeared to be

the best predictor of clients' satisfaction with their

counselors at time.

The fourth regression included Counseling Evaluation

(CEVAL) at time as the criterion variable, and the four

Trust variables at time as the predictor variables.

Consistent with the other three regressions for this

hypothesis, this model was also significant (E(4, 59) =

10.18, p < .0001), and once again Counselor Trustworthiness

(Trust3) was the only significant predictor variable (p :

.0001). The R-square (R2 = .408) was also moderate

indicating that the Trust variables at time accounted for

approximately 41% of the total variance in clients'

satisfaction with their counselors at time. The consistent

finding that Counselor Trustworthiness was the only

significant predictor in all four regressions indicates that

clients' perceptions of their counselors as trustworthy

(Trust3) may be considered an important predictor of clients'

satisfaction with counseling and with their counselors

Results of this regression model are presented in Table 4-10.

Results of the multiple regressions for hypothesis six

revealed that only moderate amounts (R2 = .368. to .460) of










the variance in SATIS and CEVAL at time and time were

accounted for by the four Trust variables indicating that


Table 4-10

Multiple Regression
the Tst- Valriable a


for Counselino Evaluation at Time3 with
a' Tim= as 0 edTic=tor


Source DF SS MS F



Model 4 10.83 2.71 10.18 .0001

Error 59 15.69 0.27

Total 63 26.53


Variable DF Parameter Standard TI
Estimate Error


Intercept 1 1.85 0.60 3.08 .0032

Trustl-3 1 0.00 0.11 0.02 .9850

Trust2-3 1 -0.03 0.06 -0.45 .6512

Trust3-3 1 0.53 0.11 4.71 .0001

Trust4-3 1 0.03 0.09 0.35 .7302

R2 = .408 CV = 12.01 Root MSE = .516 m = 4.29

Adj. R2 = .368




collectively the Trust variables were not highly influential

predictors of clients' satisfaction with their counseling

(SATIS) and with their counselors (CEVAL). However, the

Pearson correlations revealed several significant positive

correlations between the Trust variables and the satisfaction


t-h= T-t- V-JAIles At- Time3 as P Le__ginrg










measures (SATIS and CEVAL). Counselor Trustworthiness

(Trust3) in particular evidenced highly significant positive

correlations with SATIS and CEVAL at both time and time (r

= .60 to X = .67, p : .0001 ), indicating that as clients'

perceptions of their counselors as trustworthy (Trust3)

increased, so did clients' satisfaction with their counseling

(SATIS) and counselors (CEVAL). Furthermore, Trust3 was the

only significant predictor in all four regressions,

suggesting that Counselor Trustworthiness may in fact be a

good predictor of clients' satisfaction with their counselors

and counseling The results of the analyses, therefore,

provide partial support for hypothesis six.

Hypothesis seven stated that compared to clients who

plan a mutual termination with their counselors, clients who

drop out of counseling will evidence lower expectations of

counseling (Personal Commitment, Facilitative Conditions,

Counselor Expertise, Nurturance) and lower levels of

perceived trust/trustworthiness (Client Trustworthiness,

Client Trust, Counselor Trustworthiness, Counselor Trust)

prior to beginning counseling (timel). In addition, dropouts

will have less positive evaluations of counseling (Client

Satisfaction) and their counselors (Counseling Evaluation)

than mutual terminators.

As originally stated, hypothesis seven required only

analyzing the comparison between mutual terminators and

dropouts at timel (prior to beginning counseling) for the

expectation factors and the trust variables. However, upon




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EFWEAAXKB_KE247S INGEST_TIME 2014-04-09T15:52:01Z PACKAGE AA00013577_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES