Citation
The relationship between response preferences and counselor effectiveness ratings of counselor trainees

Material Information

Title:
The relationship between response preferences and counselor effectiveness ratings of counselor trainees
Creator:
Bradley, Sidney Leon, 1942-
Publication Date:
Language:
English
Physical Description:
ix, 129 leaves : ; 28 cm.

Subjects

Subjects / Keywords:
Career counseling ( jstor )
Counselor training ( jstor )
Educational counseling ( jstor )
Educational evaluation ( jstor )
Empathy ( jstor )
Personnel evaluation ( jstor )
Psychological counseling ( jstor )
Psychotherapy ( jstor )
School counseling ( jstor )
School counselors ( jstor )
Dissertations, Academic -- Education -- UF
Education thesis Ph. D
Educational tests and measurements ( lcsh )
Student counselors -- Training of ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Includes bibliographical references (leaves 110-127).
Additional Physical Form:
Also available online.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Sidney Leon Bradley.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
022479184 ( ALEPH )
04536047 ( OCLC )

Downloads

This item has the following downloads:


Full Text









THE RELATIONSHIP BETWEEN RESPONSE PREFERENCES
AND COUNSELOR EFFECTIVENESS RATINGS
OF CC'UJ',ELOR TRAINEES











By

SIDNEY LEON BRADLEY


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




UNIVERSITY OF FLORIDA


1976





































Dedicated to

Barbara Sherril Il Bradley















ACKNOWLE I .IIEU TS


The author wishes to express sincere appreciation to the following

people who assisted in this study:

Dr. Joseph Wittmer, the chairman of the committee, whose leader-

ship, experience, and precious time were invaluable assets.

Dr. Larry C. Loesch whose expertise in research design and statis-

tical analysis was essential to this study and deeply appreciated.

Dr. Paul G. Schauble whose sensitivity, encouragement, and insight

were major contributions.

Dr. James L. Lister whose instrument provided the primary focus of

this study and who originally served as the writer's committee chairman.

Ms Barbara Rucker whose skills contributed both to the computer

analysis and to the splendid job of typing this study.

Barbara, my wife, whose love, patience, and understanding were

primary sources of inner strength during my entire graduate school

experience. It is to her that this study is appreciatively and affec-

tionately dedicated.















TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS . . . .. ...... . .. iii

TABLE OF CONTENTS . . . . . .. iv

LIST OF TABLES . . . . . . vi

ABSTRACT . . . . . . .. vii

CHAPTER I INTRODUCTION I

Purpose of the Study . . . . .. .I
Rationale for the Study . . . . I
Statement of the Problem . . . . 4
Hypotheses . . . . . . 5
Definition of Terms . . . . . 5

CHAPTER II REVIEW OF RELATED RESEARCH 7

Approaches to Counselor Effectiveness . . . 7
The Speculative-Philosophical Model . . 8
The Psychometric Model . . . . 10
The Criteria Crisis in Effectiveness Model Building 13
The Communications Analysis Model . . ... 17
The Perceptual Model . . . . . 19
The Carkhuff-Truax Therapeutic Conditions Model ..... 23
Summary . . . . . .. 27
Counselor Effectiveness Ratings . . . .. 28
Supervisor Ratings . . . . . 30
Client Ratings . . . . . 31
Summary . . . . . . 34

CHAPTER III METHODOLOGY 36

Description of the Sample . . . . .. 36
Data Collection . . . . . . 38
Limitations . . . . . . 41
Instrumentation . . . . . . 42
Helper Response Preference Inventory (HRPI) . .. 42
Barrett-Lennard Relationship Inventory (BLRI) . .. 46
Counselor Evaluation Rating Scale (CERS) . .. 53
Analysis of Data . . . . . . 57








CHAPTER IV RESULTS OF THE STUDY


Introduction ........... ............................ .. 60
Descriptive Data of the Sample . . . .. 60
The Counselor Trainees . . . . .. 60
Supervisors . . . . . .. 63
Clients . . . . . . 63
Client Evaluation Scores on the Counselor Trainees . 63
Supervisor Evaluation Scores on the Counselor Trainees 64
Scores of the Counselor Trainees on the Predictor
Instrument .... .......... . . . .. 65
Relationships among the Three Instruments . . .. 66
Testing the Hypotheses . . . . . 72
Test of the First Major Hypothesis . . .. 72
Test of the Second Major Hypothesis . . .. 74

CHAPTER V DISCUSSION AND SUrIARY

Summary of the Results . . . . . 80
Practical Limitations . ..................... 81
Interpretation and Discussion of Findings . . 83
Application . . . . . . 85
Suggestions for Further Research . . . .. 87
Conclusions of the Investigation . . . .. 88

APPENDICES 90

A Counselor Trainee Demographic Data Sheet . .. 91
B Cover Letter to the Client . . . .. 92
C Relationship Inventory Form 0S-M-64 . . 93
D Relationship Inventory Form OS-F-64 . . .. 96
E Counselor Evaluation Rating Scale . . .. 99
F Helper Response Preference Inventory . . .. 101

BIBLIOGRAPHY.. . . . . . . 110

BIOGRAPHICAL SKETCH ... . . . . . .128















LIST OF TABLES


Table Page

I Split-Half Reliabilities for Three Samples: Graduate Level
Counselor Trainees, Paraprofessional Counselor Trainees
in a Community College, and Paraprofessional Disaster
Relief Counselor Trainees . . . .. 44

2 Factor Labels and Factor Scores for the Helper Response
Preference Inventory . . . . . 46

3 Test-Retest Reliability of the Subtests of the 64-Item
Relationship Inventory in Several Studies . .. 49

4 Demographic Data on the Participating Counselor Trainee
Subjects . . . . . . 62

5 Means, Standard Deviations, and Ranges of Scores on the
Barrett-Lennard Relationship Inventory (N=46) . .. 64

6 Means, Standard Deviations, and Ranges of Scores on the
Counselor Evaluation Rating Scale (N=46) . . .. 65

7 Means, Standard Deviations, and Ranges of Scores on the
Helper Response Preference Inventory (N=46) . ... 66

8 Intercorrelation Matrix of All Variables of the Barrett-
Lennard Relationship Inventory, the Counselor Evaluation
Rating Scale, and the Helper Response Preference
Inventory . . . . . . 67

9 Mean Correlation of Each Helper Response Preference
Inventory Factor with AllI of the Scores of the Two
Criterion Instruments . . . . .. 71

10 Barrett-Lennard Relationship Inventory Predictors . .. 73

II Counselor Evaluation Rating Scale Predictors . .. 75

12 Regression Equation for the Barrett-Lennard Relationship
Inventory Predictors . . . . . 77

13 Regression Equation for the Counselor Evaluation Rating
Scale Predictors . . . . . 78














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


THE RELATIONSHIP BETWEEN RESPONSE PP1FIFREJCES
AND COUNSELOR EFFECTIVENESS RATINGS
OF COUNSELOR TRAINEES

By

Sidney Leon Bradley

December, 1976

Chairman: Joseph Wittmer
Major Department: Counselor Education

This study investigated the relationship between response preferences

of counselor trainees and effectiveness ratings by their supervisors and

clients. A total of forty-six counselor trainees, enrolled in a graduate-

level practicum or internship, participated as subjects in this research.

Each counselor trainee completed the predictor instrument, the Helper

Response Preference Inventory (HRPI), which yields ten factor-analyzed

sub-scales. Client perceptions of the relationship were obtained by

having the trainee rated by a chosen client on the Barrett-Lennard Rela-

tionship Inventory (BLRI). The supervisor criterion instrument was the

Counselor Evaluation Rating Scale (CERS) on which each trainee was

evaluated by an individual supervisor.

It was hypothesized that a relationship would be found between

response preferences on the HRPI and client evaluations on the BLRI.

It was also hypothesized that a relationship would be found between HRPI

responses and supervisor evaluations on the CERS. The forward stepwise








regression technique was used to analyze the data. Four demographic

variables: age, graduate grade point average (GPA), number of graduate

quarters completed, and number of quarters of supervised counseling

practice completed, were also entered into the regression equations.

The best predictor of the BLRI Total score was grade point average

followed by responses that were premature judgments. The BLRI Level of

Regard subscore was predicted best by frank feedback responses. The

second best predictor was GPA followed by empathic responses and then

responses from an external frame of reference. The BLRI Empathy Level

subscore was predicted best by GPA, but the scale was also predicted by

premature judgments. For the BLRI Unconditionality of Regard subscore

GPA was the strongest predictor, but frank feedback was a significant

negative predictor. Only GPA predicted the BLRI Congruence subscore.

Other than GPA the demographic variables were not significant predictors.

The only statistically significant predictor of the CERS Total score

was the set of responses which used clarification primarily. The

clarification response factor also was the best predictor of the

Counseling subscore of the CERS although the scale was also predicted

by responses that validate or affirm the counselee's identity. There

was no predictor of the CERS Supervision subscore. No demographic

variable predicted any of the CERS scores.

It was noted that no HRPI factor predicted successfully both BLRI

and CERS scores. Also the canonical correlation between any two of the

three instruments in the study was not found to be significant. Inter-

correlations between the instruments were also generally low. Never-

theless, neither hypothesis, that a relationship exists between the HRPI

and each of the two criterion instruments, could be rejected except in

part.


viii








The investigator concluded that the HRPI predicts both client and

supervisor evaluation scores on the criterion instruments in differing

ways. Self-disclosing responses appeared to predict higher client

ratings while clarifying responses tended to predict higher supervisor

ratings in this study.
















CHAPTER I

INTRODUCTION

Purpose of the Study


The measurement of counseling effectiveness and prediction of

counseling success are both central concerns to the counseling pro-

fession in the era of accountability. Efficiency and professional

responsibility dictate the need for valid ways to evaluate competence

and to differentiate levels of effectiveness. This study has explored

the relationship between the effectiveness ratings of counselor trainees

and their characteristic ways of responding to clients as a possible

method of evaluating effectiveness. More generally, this study has

investigated the theory that a counselor's communication within the

helping relationship is related to his level of effectiveness as a

helping person. This study also attempted to specify measurable compo-

nents of communication which distinguished more effective counselor

trainees in a graduate level training program from less effective

counselor trainees.


Rationale for the Study

Research that leads to identification of effective counseling is

clearly relevant to the process of counselor selection (ACES, 1967),

evaluation (Stablein, 1962), and prediction of future counseling behav-

ior (Walton & Sweeney, 1969). Nevertheless, investigations which have

attempted to specify and predict effective counseling have frequently

I








been disappointing and inconclusive. One result has been that admission

criteria to counselor training has been severely criticized for showing

little relationship to effective counseling (Rapaport, 1968; Thoresen,

1969). These indictments appear warranted in light of the research which

suggests that counseling can indeed be "for better or for worse" (Car-

khuff, 1966; Lister, 1970). A mounting body of research indicates that

unless a counselor can embody personal facilitative qualities such as

empathy, warmth, genuineness, and concreteness, he may adversely affect

the client (Carkhuff & Berenson, 1967). The profession is challenged

by these findings to devise and employ more effective selection criteria

to increase the probability of a counselor making a beneficial rather

than a detrimental impact on the client (Thoresen, 1969).

Intellectual attributes have been used extensively to screen

candidates for counselor training programs. Unfortunately, while

ability measures may predict graduate grade point average with reason-

able accuracy, neither aptitude test scores nor grade point averages

yield a significant correlation with counseling effectiveness (Car-

khuff, Piaget, & Pierce, 1968; Bergin & Solomon, 1968). The Graduate

Record Examination, although widely used in graduate counselor educa-

tion programs for selection purposes, appears to provide little meaning-

ful predictive validity (Wittmer & Lister, 1971).

Since 1963, when the American Personnel and Guidance Association

encouraged a shift to the evaluation of personality variables in

counselor candidates, a vast array of personality correlates with

counseling effectiveness have been examined. Illustrative of this

attention to the personality of the counselor were the reviews of

Burnett (1954),Stoughton (1957), Hill and Green (1960), Stripling and







Lister (1963), Froehlich (1951), Cash and Munger (1966), and Whiteley

(1969). Again, the results have been conflictual and inconclusive.

In his review of research in this area, C. H. Patterson concluded that

the data were sporadic, often poorly integrated with the basic issue

questions, and generally inadequate (1967).

An entirely different area of investigation began in 1950 with the

publication of the book Therapeutic Counseling by E. H. Porter. His

book presented the thesis that the counselor's response pattern tends

to enhance or impede the flow of helpful communication. He contended

that the responses a counselor selects from his broad range of options

have implications for his ability to be an effective helping person.

His assertion seemed to imply that the primary index of a counselor's

ability to interact helpfully was the skill with which he could communi-

cate with another human being. This appears to hold today.

In a similar vein, an entire school of counseling and psycho-

therapy has developed around communication as the central concept in

human interaction (Bateson et al., 1956; Jackson, 1965; Haley, 1963;

Satir, 1967). While communication involves much more than simple

verbal exchanges, the communication theorists emphasize the importance

of the "report" or "message" component. Their belief is that largely

through verbal responses to himself, or "reflected appraisals" as

Sullivan (1953) called them, a person constructs his self concept, his

conception of the outside world including percepts of other persons,

and his sense of identity based upon how he personally fits into that

outside world. These theoretical constructs have been supported in

part by the process research on counseling which has linked at least

four variables to beneficial client outcome (Carkhuff & Berenson,








1967). These variables, empathy, warmth, genuineness, and concrete

specificity appear to be integrally related to, and generally are

measured by, the degree of sensitive, appropriate responsiveness

demonstrated by the person in the counselor role.

The question which naturally emerged from the above was whether

pure communicational response style was sufficient to identify and pre-

dict effective counseling. For example, if Porter's assumption that

understanding, interpretive, etc. responses facilitate beneficial

client change were empirically valid, then perhaps an investigation of

response preferences could reveal one crucial factor in effective coun-

seling. In fact, it might be possible to provide stimuli similar to

the counseling situation and test on paper how a counselor candidate

or trainee would prefer to respond in some future counseling relation-

ship. This study explored whether those assumptions were warranted in

reference to one particular instrument for measuring response preferences.


Statement of the Problem

This study has investigated whether or not a measurable relation-

ship exists between the effectiveness ratings of a counselor in training

and certain interpersonal communicational modes. The question was

examined utilizing two discrete criteria. One criterion related

preferred verbal responses to client perceptions of the relationship

shared with the counselor trainee. The second criterion explored the

relationship of counselor trainee response preferences to the evaluation

of effectiveness made by the counselor trainee's individual supervisor.

More specifically, the questions answered by this study are:

I. Does a counselor trainee's response preferences on a printed








counseling-simulation instrument have a significant relation-

ship to clients' perceptions of a beneficial relationship?

2. Does a counselor trainee's response preferences on a printed

counseling-simulation instrument have a significant relation-

ship to supervisors' evaluations of counseling effectiveness?


Hypotheses

The general hypothesis in this research was that a relationship

would be found between the response preferences of counselor trainees

on the experimental instrument and two separate measures of counseling

effectiveness; namely, supervisor ratings and client perceptions of the

relationship. The following sub-hypotheses were tested:

HOI There is a relationship between counselor trainee response

preferences and client evaluations of their counseling

effectiveness.

HO2 There is a relation-hip between counselor trainee response

preferences and supeor isor evaluations of their counseling

effect iveness.


Definition of Terms

In this study the follo inpc- definitions of relevant terms apply:

I. Counselor Response Preference The choice of a verbal reply to a

client's stimulus statement. In this study the helpee statement on

each item of the Helper Reso,-onse Preference Inventory was the

stimulus. The counselor trainee was asked to choose one of two

helper responses for his ro,:ly. Since the format was forced-choice,

the range of options was I i"-ited to only two response possibilities.

2. Counselor Trainee Graduate students in Counselor Education








pursuing either a specialist or doctoral degree at the University

of Florida. Each trainee was currently engaged in either a super-

vised practicum or internship in counseling at the time of data

collection.

3. Client Ratings The perception by a client of the helpful quality

of the relationship with the counselor trainee. Operationally, the

client rated the counselor trainee on the Barret-Lennard Relation-

ship Inventory. Higher scores were assumed to be indicating of a

more helpful relationship in the perception of the client.

4. Supervisor Evaluations The perception of the counselor trainee's

effectiveness by his individual supervisor. The ratings were

operationalized as scores on the Counselor Evaluation Rating Scale.

Higher scores indicated the supervisor perceived the counselor

trainee to be relatively more effective as a helping person.

5. Counseling Effectiveness The ability to a counselor trainee to

beneficially affect a client. For the purposes of this study,

counseling effectiveness was operationally defined in terms of

ratings of the counselor-in-training by clients and supervisors.














CHAPTER II

REVIEW OF RELATED RESEARCH


The review of literature focuses on the following areas: I) the

varied approaches to conceptualization and measurement of counselor

effectiveness, and 2) the application of supervisors' and clients'

ratings to counselor effectiveness research.


Approaches to Counselor Effectiveness

The pursuit of a consistently viable way to assess the effective

counselor has often led to confusing and contradictory results. Quite

early, the sentiment had crystallized, and was later traced historically

by Polmantier (1966), that the profession needed to recruit counselor

candidates with personal qualities which would enhance their success as

counselors. Subsequently, the process of identification of those

essential personal qualities has met with mixed results. In this review,

five primary approaches to the search for a definition of counselor

effectiveness are explored: The speculative-philosophical approach, the

psychometric approach, the communication analysis approach, the percep-

tual approach, and the core dimensional approach derived from process

research. After the first two approaches are presented, the question of

the criterion problem in effectiveness research is discussed to intro-

duce the consideration of the final three approaches to counselor

effectiveness research.







The Speculative-Philosophical Model

The most primitive approach to establishing the components of

effective counseling began with the emergence of serious attempts to

conceptualize the necessary ingredients. In the widely quoted Fiedler

(1950) study it was concluded that experienced counselors, regardless

of frame of reference, agreed on the fundamental characteristics of an

ideal counselor-client relationship. The counselors concurred that the

ideal therapeutic relationship involves I) an empathic relationship,

2) rapport between client and counselor, 3) the counselor staying with

the client's problems, 4) the freedom to self-disclose, 5) mutual trust

and confidence, 6) excellent rapport, 7) active client participation,

8) freedom to make choices by client, 9) acceptance of client feelings

as normal and understandable, 10) a tolerant atmosphere, and II) a

feeling of being understood on the client's part.

Concurrent with the analysis of the ideal relationship was a

growing interest in those personal qualities of the counselor which

engender and enhance this therapeutic relationship. In their review of

this early literature, Barry and Wolf (1958) concluded that the

philosophical approach in this era consisted primarily of cataloguing

assorted virtues to be sought in prospective counselors. Graves (1944)

specified integrity, vitality, judgment, industriousness, high personal

standards, adaptability, experience and training. Cox (1945) added such

qualities as fairness, sincerity, good character and wholesome philosophy.

In addition to the above, Bowler and Dawson (1948) emphasized maturity,

respect, resourcefulness, reliability, a sense of humor, and an ability

to listen and keep confidences. Then in 1949 the National Vocational

Guidance Association proclaimed that counselors, ideally, were interested








in people, patient, sensitive to others, emotionally stable, objective,

respectful of facts, and trusted by others.

Hamrin and Paulsen in 1950 attempted to introduce a slightly more

scientific approach. They conducted a field study of 91 counselors to

arrive at a consensus of desirable traits which they believed facilitated

counseling. In descending order of frequency they listed I) under-

standing, 2) sympathetic attitude, 3) friendliness, 4) sense of humor,

5) stability, 6) patience, 7) objectivity, 8) sincerity, 9) tact, 10)

fairness, II) tolerance, 12) neatness, 13) calmness, 14) broadminded-

ness, 15) kindliness, 16) pleasantness, 17) social intelligence, and

18) poise.

By 1960, however, Hill and Green reported that the enthusiastic

search for the essential personality pattern of a counselor had all but

dwindled away. The focus by then had shifted from speculative lists of

character traits to examining the total personality of the counselor

and how that personality impacts upon the counseling relationship. This

transition naturally arose out of a growing disenchantment with the

subjectivity of the philosophical-speculative approach (Cottle, 1953;

Hill and Green, 1960; Jones, 1951). While no massive revolution in the

understanding of effective counseling resulted from the philosophical

pursuit of ideal counselor characteristics, such outgrowths as Rogers'

formulation (1957) concerning empathy, genuiness, and unconditional

regard proved to be of considerable heuristic value. Cottle (1953)

recognized the helpfulness of the trait listings but criticized them

because the lists (I) were merely opinions, (2) failed to distinguish

counselors from related professionals, (3) varied widely, and (4)

failed to specify which patterns or interrelations of traits were








important. Consequently, the pursuit of measurable personality

variables of effective counselors began in earnest using more objective

measures.


The Psychometric Model

The use of psychometric test data to link personal variables with

counseling effectiveness has been a popular, broad-based movement.

Actually a number of researchers have been able to effectively discrimi-

nate quality levels in counseling using personality measures. However,

in general, the differentiations between more and less effective

counselors have not succeeded in identifying a dependable instrument

for screening. The principal reasons why such an instrument has not

been found are that positive findings are seldom replicated and results

add up to be in consistent and inconclusive. In general, the psycho-

metric approach has encompassed personality, interest, and attitude

measures of counselor effectiveness. To simplify the proliferation of

research in this area, this review will examine several popular measures,

one at a time.

The Edwards Personal Preference Schedule (EPPS) has been widely

used as a personality instrument to discriminate effective from less

effective counselors. However, the results overall have been per-

plexing. Stefflre, King, and Leafgren (1962) used the Edwards Personal

Preference Schedule to examine participants in a one semester institute.

They found "most chosen" counselors, on the basis of potential, scored

higher on "deference" and "order" and lower on "aggression" and "abase-

ment" than the "least chosen" counselors. On the other hand, the

counselors judged to be more effective in the study by Truax, Silber,







and Wargo (1966) were found to have higher scores on "change" and

"autonomy" and lower scores on "order." Ohlsen (1967), also using the

Edwards Personal Preference Schedule, measured two separate sample groups

of counselors. In the first group the most effective counselors scored

higher than the least effective on "succorance" and lower on "order,"

"deference," and "consistency." In contrast, Ohlsen's most effective

counselors in the second group scored higher than the least effective

on "intraception" and lower on "dominance" and "aggression." Demos and

Zuwaylif (1966) found the EPPS to be the only one of three psychometric

measures to discriminate accurately among 30 secondary school counselors

in an NDEA institute based upon "multiple objective and subjective

criteria." Their most successful counselors scored significantly higher

on "nurturance" and "affiliation" while the least successful group

scored significantly higher on "autonomy," "abasement," and "aggression."

It is apparent that despite the similarities overall, the general

pattern of results is much to inconsistent to permit the broad scale

use of the EPPS as a reliable predictor of effective counseling.

Another standardized personality instrument which has been applied

in attempts to measure counselor effectiveness is the 16 Personality

Factor Profile (16 PF). In 1968 a regression equation was derived by

McClain for predicting counselor effectiveness from 16 PF scores. The

limited use of this prediction equation may be related to the generally

significant but inconsistent results found in early research attempts

to use the 16 PF to differentiate levels of effectiveness (Donnan,

Harlan and Thompson, 1969; McClain, 1968; Myrick, Kelly, & Wittmer,

1971; and Wittmer & Lister, 1971).

The Rokeach Dogmatism Scale has been used with considerable promise








in this area. Walton and Sweeney (1969) state that in their review

they found that "results of work utilizing this instrument have rather

consistently supported its capability for distinguishing between coun-

selors on opposite ends of the effectiveness continuum" (p. 34).

Stefflre, King, and Leafgren (1962) found, among counselor trainees

selected as having more counseling potential by their peers in an 'L'EA

Guidance and Counseling Institute, significantly less dogmatism than,

among those trainees not chosen. Cahoon, in 1962, also reported that he

had found significantly less dogmatism in superior counselors. These

findings have been supported by Kemp's finding (1962) that low scores in

dogmatism on the Dogmatism Scale were related to openness in an actual

counseling situation. Still further corroboration came from the conclu-

sion by Russo, Kelz, and Hudson (1964) that openmindedness is indeed an

important quality in a counselor after they found 12 items on the

Dogmatism Scale which distinguished significantly between counselors

judged most and least effective. Furthermore, in studies that did not

use the Dogmatism Scale but measured a closely related factor, tolerance

for ambiguity, both Brams (1961) and McDaniel (1967) found a relation-

ship between this factor and high ratings of effectiveness.

Virtually a myriad of other personality measures have been tried in

attempts to isolate personality patterns which appear to coincide with

effective counseling. Some illustrative efforts have included score

profiles from the Miller Analogies Test and the MMPI (Abeles, 1958),

the MMPI alone (Brams, 1961; Foley & Proff, 1965), the Strong Vocational

Interest Blank (Stefflre, King, & Leafgren, 1962), the Kuder Preference

Record (Blocher, 1963), the Wisconsin Relationship Orientation Scale

(Wasson, 1965), the Allport-Vernon-Lindzey Scale of Values (Demos &








Zuwaylif, 1966), the Self-Description Checklist (Wicas & Mahan, 1966),

the California Personality Inventory and the 'IMPI (Schroeder & Dowse,

1968), and the Interpersonal Check List (Eberlein & Park, 1971). While

some of these studies have reported promising avenues for reliable

prediction, frequently the attempts to replicate have either yielded

conflictual findings or qualified the original results in ways that

diminish confidence in the ultimate dependability of the instrument.

An excellent illustration of this frustrating process has been the

attempts to differentiate the measurable characteristics of A and B

type therapists as first described by Whitehorn and Betz (1954, 1960).

They found that type A therapists who established warm, personal rela-

tionships had substantially better success rates with schizophrenics

than the more passively detached and interpretive type B therapists.

Furthermore, they found that 23 items on the Strong Vocational Interest

Blank distinguished the two therapist types. However, McNair, Callahan,

and Lorr (1962) found that with neurotic patients, and psychologists as

therapists, the results were reversed and the type B therapists were

more successful. Carson et al. (1964) concluded from their research

that the type A therapist works best with distrusting clients and type

B with trusting clients, in an attempt to reconcile the discrepancies.

Kemp (1966) introduced still more controversy by concluding from his

research that both types of therapists are most capable of helping those

clients who produce the most discomfort in themselves. It is readily

apparent that the findings are intriguing but decidedly inconclusive.


The Criteria Crisis in Effectiveness Model Building

The disappointing attempts to link measurable personality variables








to counseling effectiveness is mystifying. It is enigmatic why results

overall have been either contradictory or confusing. One criticism of

this entire approach came very early from Stern, Stein, and Bloom (1956).

They felt the inconsistency eventuated from the inadequacy of the trait

factor philosophy in which the approach was rooted. They called for

attempts to identify meaningful personality traits related to counseling

behavior only after predictions have been derived from well-articulated

theories.

Other observers take a different perspective and view the problem

as an artifact of the "effectiveness" criteria invoked. The profession

of counseling has not attained a consensus on the goals of counseling,

the differential counseling methods appropriate to specific problems,

or standard indicators of client progress. Therefore, the elusive con-

cept of "counselor effectiveness" may well be relative not only to the

uniqueness of the counselor, but also to the setting, the problem, and

the client (Stefflre, King, & Leafgren, 1962).

Stefflre, King, and Leafgren (1962) pointed out that we are unable,

at our present knowledge level, to reach consensus about how to identify

or even define the effective counselor. Nevertheless, societal pressure

urges the profession to educate "as if" the "effective counselor" was a

definable entity. Generally, ratings of supervisors or judgment by pro-

fessional practitioners has been used to differentiate levels of skill,

but client ratings have gained in popularity and are explored in depth

in the next section of this chapter. However, the ratings utilized have

proliferated in effectiveness studies to include peer judgments, self-

ratings, etc., which has added to the profusion of confusion.

Researchers have attempted at times to look at behavioral criteria








or "unobtrusive measures" with a variety of behaviors being studied.

Evaluation of counseling success has varied from dropout rate to atten-

dance and grades to the proportion of time spent talking by the client.

A serious question, however, has been raised by Hill (1960) regarding

the validity of such outcome criteria. In reality such behavioral

criteria are responsive to change effects from a host of uncontrollable

and often unidentifiable variables.

In their 1960 review of research, Hill and Green traced a large

portion of the difficulty in finding a consensual definition of effec-

tiveness to the complexities of the settings within which counselors

function. They cited the literature to support their contention that

the issues of selection and effectiveness are often confused. Large

variations in roles and relationships exist even in one counselor's work

(APA, 1958; Hoffman, 1959), and across jobs the diversity of skills and

attitudes needed by counselors becomes staggering. There is growing

awareness that often, as Hill and Green (1960) assert, the counselor may

be called upon to fulfill responsibilities which the profession has

neither embraced nor prepared the counselor to do (Purcell, 1957;

Wellman, 1957). Today the role and function of the counselor appears to

be diverging even more from a precisely defined model which is amenable

to unambiguous evaluation.

Beyond the lack of agreement as to what is the preferred model of

effective treatment for a specific clientele, lies the diversity of

opinion about the different schools of therapy. In one study Sunderland

and Barker (1962) found that experienced therapists from Freudian,

Sullivanism, and Rogerian schools exhibited different behaviors, but

precisely how the specific counseling approach affects the effectiveness








of the counselor in the wide range of possible settings, c ">-s, and

problems is obscure.

Still we have not exhausted the variables that may cc-.--_-. an

attempt to define and measure an effective counseling rel--:-hip.

Some variables that have gained attention are: the attit-:''-' and

value system of the therapist (Combs & Soper, 1963; White- -- Betz,

1954); demographic variables (Bailey, Warshaw, & Eichler, G> ; the

experience of the therapist (Cartwright & Vogel, 1960); t-e e:'-ee of

disturbance of the patient (Stone, Frank, Nash, & Imber, I ,; and the

client's felt need to change (Cartwright & Lerner, 1963). 3:. ;ously

this list is not comprehensive, but it clearly illustrates --- nost

of variables which have been found to influence therapeutic e-ctive-

ness. If all the variables which it is possible to study ..e-e added to

the list of those which have been examined, the resulting v-:3site

would likely be astronomical.

For some the complexity of teasing effectiveness out i -s laby-

rinth of variables is felt to be overwhelming. A few conc.-e -nat

counseling effectiveness research is so elusive as to ever :e :eiond

critical inquiry (Thoresen, 1969). Some writers feel the -"ec ical

foundations of counseling have been shaken (Ford & Urban, and

others fear the super-structure of the profession is dange-c-z close

to collapse into chaotic confusion and revolution (Colby, :-Z:
boltz, 1966; Rogers, 1963). Simultaneous with this express: :mncern

is the growing pressure for the profession to be increasing: :count-

able. The ground swell of public demand calls for the coups= -: pro-

fession to demonstrate selection, evaluation, and educatio- :--:edures

which are derived from and justified by a substantial body -- :-:irical

research on effectiveness (Thoresen, 1969).








Perhaps the most viable alternative is for the profession to move

more toward the "systems approach" to conceptualizing and evaluating

counseling. A key concept is that the system is more than the sum of

its parts. For example, the counseling encounter is a total relation-

ship built upon two distinct but temporally engaged personalities

attempting to bridge their separateness via a communications process.

The crucial variables may be the unique perceptual worlds of the coun-

selor and client, the communication that flows between them, and the

emotional tone of the atmosphere within which they transact. Perhaps

this triad offers a viable and comprehensive approach to the enormously

complex give-and-take of the counseling process. Since the variables

under investigation in this study relate specifically to communication,

it is important to consider the relevance of this factor to counseling.


The Communications Analysis Model

Fiedler (1950, 1951) pointed out the crucial significance of the

relationship to counseling. He reported that experts excel non-experts

in their ability to communicate and understand their patients. Porter

(1950) went a step further and hypothesized that the counselor's style

of preferred responses sets the tone of the interview and defines the

counselor's capacity to be helpful. In Porter's instruments, the Porter

Test of Counselor Attitudes and the Porter Interview Analysis Scale, he

categorized responses as understanding, supportive, interpretive,

probing and evaluative. A number of studies of trainees in counselor

education programs have revealed that with exposure to training and

supervision trainees become more understanding and less evaluative,

supportive, and probing (Demos & Zuwaylif, 1963; Kassera & Sease, 1970;









Munger & Johnson, 1960). A paradoxical finding, however, has been dis-

covered among rehabilitation counseling trainees where preferences

are decidedly in favor of responses that inform, question, probe and

diagnose (Davis & O'Conner, 1974; Wittmer & Lister, 1971).

In his doctoral dissertation Chapline (1964) related level of cog-

nitive complexity in counselors to preferences for understanding or

evaluative responses. He administered the Construct Repertory Test to

his sample of 163 graduate students in counseling and guidance from the

departments of psychology and education to measure cognitive complexity.

In addition to gathering demographic data, he had the subjects respond

in writing to an experimental tape recording of client statements.

Using Porter's original categories (1950), he found that subjects who

measured high on cognitive complexity were also high on evaluative

responses while the low cognitive complexity scorers were high on

understanding responses. The education students tended to prefer

evaluative responses and the psychology students leaned toward under-

standing responses at a higher-than-chance level. On the other hand,

response preferences were unrelated to sex, age, level of training,

previous personal counseling, teaching experience, or even experience in

counseling. Freedman, Antenem, and Lister (1967) in a related study

found a strong predictable relationship between verbal response patterns,

as measured by Porter's Scales, and counselor personality characteris-

tics on the California Psychological Inventory. The response-analysis

of their 15-minute roleplayed interviews was suggested as a useful

procedure for student selection and progress evaluation.

Grigg and Goodstein (1959) found that clients who saw their coun-

selors as verbally active were much more likely to report a favorable










outcome than those who saw their counselors as passive listeners. In

his 1961 report Brams emphasized the strong tie between the counselor's

personality and his ability to communicate with clients, and Pallone

and Grande (1965) even concluded that the counselor's verbal mode sig-

nificantly influences the amount of the client's problem-relevant

communication. Clearly the counselor's communication style is funda-

mental to the relationship established with the client. However, the

extent of this influence has not been established. This study, in part,

will be an attempt to evaluate the impact of verbal response preference

to a multi-factor measure of counseling effectiveness.

In the belief that the counselor's communicational mode is a direct

expression of one's unique perceptual world, and that in turn one's

communicational style crucially shapes the atmosphere of the counseling

relationship; the focus of this review now turns to the last two

approaches to counseling effectiveness relevant to this study: the

perceptual approach and the core therapeutic conditions approach as

epitomized in the writings of Carkhuff.


The Perceptual Model

A much different approach to the question of counselor effective-

ness has been introduced by perceptual psychology. According to this

approach (Combs & Snygg, 1959), the internal perceptions and perceptual

organization of the individual provide a meaningful perspective for the

examination of effectiveness in human helping. The assumption is made

that counselors who have "learned to use themselves as effective instru-

ments in the production of helping relationships can be distinguished

from those who are ineffective on the basis of their characteristic









perceptual organizations"(Combs & Snygg, 1959, p. 14). The perceptual

approach then may be contrasted with more behavioristic approaches which

operationalize the criteria of effective helping in terms of specifiable

behaviors.

According to the perceptual approach, a man's behavior is a func-

tion of his "perceptual field," By perceptual field is understood the

individual's unique perceptions of himself and the world in which he

lives. Hence, the field consists of what the person perceives at a

given point in time and the meaning or significance he attaches to that

perceptual experience (Combs, Avila, & Purkey, 1971). Behavior is seen,

therefore, as a direct consequence of the individual's field of meanings

existing for him at the moment of behaving. Since the perceptual

psychologist is concerned with internal impression rather than external

expressions, the internal frame of reference is preeminent. Perhaps it

would not be an overstatement to say that the perceptual theorist

believes that a counselor is effective not because of his specific

behaviors but rather because of the internal perceptions behind those

behaviors.

Walton and Sweeney (1969) cite the study of counselor perceptions

by Combs and Soper (1963) as one of the more promising approaches to the

measurement of effectiveness. Using a critical incidents in human

relationships format, they found moderately high positive relationships

between 12 characteristic ways of perceiving and effectiveness as a

counselor. The effective counselor was sensitive to and concerned with

how things looked to others and was oriented to people rather than

things. He perceived others as able rather than unable, dependable

rather than undependable, friendly rather than unfriendly, and worthy








rather than unworthy. Moreover, he perceived himself as being identi-

fied with people rather than apart from people, as personally adequate

rather than wanting, and as self-revealing rather than self-concealing.

These dichotomous characteristics which Combs and Soper studied

grew out of a graduate seminar held in 1959 at the University of Florida

(Combs, 1961). The purpose of the seminar was to investigate effective

helping relationships from the internal point of view. The effort to

identify perceptual characteristics which were believed to be relevant

to the task of effective professional helping crystallized in five focal

areas (I) the general frame of reference from which the helper approaches

the situation, (2) the ways in which the helper perceived other people,

(3) the ways in which the helper perceived himself, (4) the ways in which

the helper perceived the task with which he was confronted, and (5)

the ways in which the helper perceived appropriate methods for carrying

out his purposes. The forty dichotomies issuing from that seminar are

listed below. Those perceptual dichotomies believed related to effec-

tive helpers are on the left and those related to ineffective helpers on

the right.

(I) General Frame of Reference

Internal External
Growth orientation Fencing in or controlling
Perceptual meanings Facts, events
People Things
Hopeful Despairing
Causation oriented Mechanics oriented

(2) Perceptions of Other People Sees others as:

Capable Incapable
Trustworthy Untrustworthy
Helpful Hindering
Unthreatening Threatening
Respectable No account
Worthy Unworthy








(3) Perceptions of Self Sees self as:


Identified with people
Enough
Trustworthy
Liked
Wanted
Accepted
Feels certain, sure
Feels aware
Self-revealing


Apart from people
Wanting
Not trustworthy
Not liked
Not wanted
Not accepted
Doubt
Unaware
Self-concealing


(4) The Helpinq Task and Its Problems


Purpose is:
Helping
Larger meanings
Altruistic
Understanding
Accepting
Valuing Integrity

Approach to problem is:
Positive
Open to experience
Process oriented
Relaxed
Awareness of complexity
Tolerant of ambiguity


Dominating
Smaller meanings
Narcissistic
Condemning
Rejecting
Violating integrity


Negative
Closed to experience
Ends oriented
Compulsion to change others
Oversimplification
Intolerant of ambiguity


(5) Perceptions of Methods Sees appropriate methods as:


Helping
Cooperation
Asserting
Acceptance
Permissive
Open communication
Giving
Vital


Manipulating
Competition
Appeasing
Rejecting (attacking)
Authoritarian
Closed communication
Withholding
Lifeless


(Combs, 1961, pp. 56-57)

More than twenty studies have attempted to infer the perceptual

organization of helpers using this variety of original perceptual dimen-


sions plus a few that have been identified subsequently.


Among the


more notable studies are those of priests (Benton, 1964), secondary

school teachers (Brown, 1970), counselor trainees (Combs & Soper, 1963),








junior college teachers (Dedrick, 1972), nurses (Dickman, 1967), college

professors (Doyle, 1969), elementary school teachers (Gooding, 1974),

residence assistants (Jennings, 1973), and counselors (Rotter, 1971).

The overall impression is that this substantial body of research suggests

a potentially rewarding means of identifying helpers with relatively

higher levels of helping ability.


The Carkhuff-Truax Therapeutic Conditions Model

The model of counselor effectiveness research which was popularized

by Truax and Carkhuff, and currently is often identified with the latter,

was formulated primarily out of the early research efforts of the pair

(Truax & Carkhuff, 1967). Their joint investigations were directed at

specifying the primary core of interpersonal facilitative factors which

relate to favorable outcome in counseling and therapy. After extensive

research, the two concluded that the core dimensions of accurate empathy,

nonpossessive warmth, genuineness, and concreteness account for between

one-third to one-half the variability in client outcome (Truax, 1961;

Truax & Carkhuff, 1966). Counselor trainees who offer their clients

high levels of these core dimensions have been found to consistently

elicit beneficial therapeutic gain (Berenson, Carkhuff, & Myrus, 1966;

Carkhuff & Truax, 1965; Truax & Mitchell, 1971).

The work of Truax and Carkhuff is clearly an outgrowth of the

Rogers' formulations. Truax, a student of Rogers, developed a "Tenta-

tive Scale for the Measurement of Depth of Self Exploration" along with

other scales rooted in the theoretical hypothesis of "necessary and

sufficient conditions" of therapeutic personality change proclaimed by

Rogers (1957). Much of the plethora of research summarized by Truax









and Carkhuff is based upon these scales. Carkhuff, after the two

researchers went divergent paths, modified each of Truax's therapeutic

conditions scales to incorporate five levels for each scale (Carkhuff,

1969a; Carkhuff & Berenson, 1967). The five levels each represent a

qualitative degree of therapeutic functioning, allowing quantification

of the extent to which a therapist corresponds with the standardized

criteria. At first the scales were used to rate excerpts of audio

taped counseling interviews. More recently video tapes have been the

medium under investigation (Kagan & Krathwohl, 1967), particularly in

the evaluation of the Interpersonal Process Recall (IPR) technique.

Even though the reliability and validity of the Carkhuff scales are

at times questioned (Hefele & Hurst, 1972), the summarization of the

extensive body of research (Carkhuff & Berenson, 1967) presents one of

the most persuasively supported theoretical positions in counseling

today. Carkhuff's scales are probably the most widely used rating

scales in current studies of self-exploration and the facilitative con-

ditions of empathy, warmth, genuineness, and concreteness.

A central concept in the model for effective counseling which is

presently being explicated by Carkhuff is that facilitative conditions

can not only be measured but can also be transmitted to counselor

trainees by counselor educators who themselves provide them (Carkhuff &

Truax, 1966). Since the accumulating weight of research supports the

belief that counseling can be "for better or for worse," depending

upon the presence or absence of these dimensions, the implications of

this research for counselor selection and accountability monitoring are

clearly evident. Since the factor subscales of the predictor instrument

in this research relate directly to the antecedent research of Carkhuff









and his associates, it seems appropriate to elucidate how he concep-

tualizes the integration of the facilitative dimensions into the thera-

peutic process. Two of his primary propositions and their corollaries

follow. These bold assertions are supported by research evidence in the

book Helping and Human Relations, Volume I (1969) from which they were

extracted.

Proposition I. The degree to which the helping person offers
high levels of facilitative conditions in response to the
expressions of the person seeking help is related directly
to the degree to which the person seeking help engages in
processes to constructive change or gain.

Corollary I. The degree to which the helping person offers
high levels of empathic understanding of the helpee's
world is related directly to the degree to which the helpee
is able to understand himself and others.

Corollary II. The degree to which the helping person communi-
cates high levels of respect and warmth for the helpee and
his world is related directly to the degree to which the
helpee is able to respect and direct warm feelings toward
himself and others.

Corollary III. The degree to which the helper is helpful in
guiding the exploration to specific feelings and content is
related directly to the degree to which the helpee is able
to make concrete his own problem areas.

Corollary IV. The degree to which the helper is responsively
genuine in his relationship with the helpee is related to the
degree to which the helpee is able to be responsively genuine
in his relationship with himself and others.

Proposition II. The degree to which the helping person
initiates action-oriented dimensions in a helping relation-
ship is directly related to the degree to which the person
seeking help engages in processes that lead to constructive
change or growth.

Corollary I. The degree to which the helper can be freely,
spontaneously and deeply himself, including the disclosing
of significant information about himself when appropriate,
is directly related to the degree to which the helpee is
able to be genuine and self-disclosing in appropriate
relationships.


Corollary II. The degree to which the helper actively









confronts the helpee and himself is directly related to the
degree to which the helpee is able to confront himself and
others.

Corollary III. The degree to which the helper both acts and
directs the actions of the helpee immediately in the present
to the relationship between helper and helpee is related to
the helpee's ability to act with immediacy and later to
direct the actions of others.

Corollary IV. The degree to which the helper can make con-
crete a course of constructive action is related to the
degree to which the helpee can go on to make concrete courses
of action for himself and others.(pp. 84-90)

The complete text of Carkhuff's working assumptions emphasizes

the direct personal effect upon the helpee of the helper's level

of facilitative conditions offered in the interpersonal encounter.

It follows that the degree to which the counselor in the helping role

provides high levels of empathy, warmth, respect, genuineness, and

concreteness is directly related to the internalization of these

facilitative dimensions into his own life. Moreover, the degree to

which the helper is openly and deeply himself, self-disclosing, con-

frontive of himself and his client, present with immediacy, and active

with clear, concrete, action-oriented alternatives is again directly

related to the client's capacity to translate these same facilitative

dimensions into his own life sphere.

Carkhuff has written extensively about the selection process and

its relationship to the previously stated propositions. He has also

addressed himself to the effectiveness criteria problems already pre-

sented. His position is that "the best index of a future criterion is

a previous index of that criterion" (1969b,p. 85). In general, the

counselor selection process must be distilled to identification of

specific functioning within the helper role by the prospective helper.









He suggests the following specific indices: I) an index of communica-

tion; 2) an index of trainability; 3) an index of self exploration; and

4) a critical incident index to ascertain his potential helpfulness in

the face of crisis. This study relates to the first and last of these

indices because the predictor instrument is a measure of communication

within a critical-incidents or counseling-simulation format.


Summary

Despite voluminous research with a myriad of approaches, the effec-

tive counselor remains a nebulous concept. The speculative-philosophi-

cal approach helped clarify for training purposes the ideal characteris-

tics and proved thought-provoking for a period, but the approach came

under critical fire because its subjectivity did not solidify the base

of the profession.

Empiricism, in the form of psychometric testing, came into vogue

as attention focused upon the measurable personality traits of the

counselor. There was not found, however, many impressive links between

specific personality traits and beneficial client outcome. It appeared

that a more global assessment of the total impact of the helper's

personality rather than a trait-factor approach was more relevant to the

therapeutic involvement of a truly helpful relationship.

Perceptual dimensions have attracted considerable attention in the

counseling profession, but research on this approach has been less than

prolific. Focus upon the internal frame of reference of the counselor

has yielded some meaningful links with effective helping.

More recently, the pressures within and without the profession

toward accountability have led to attempts to define counseling









effectiveness in terms of demonstrable client outcome criteria. In

response, the Carkhuff model has endeavored to derive a coherent system

of selection and training based upon their conclusions from outcome

research. Nevertheless there remain the ubiquitous problems of effec-

tiveness criteria, plausible indicators of therapeutic gain, consensual

counseling goals, and quantifying the variability in clients, settings,

and problems. While the ultimate goal has not been attained, the

impression is that greater definitional and procedural clarity have

evolved from the continual grappling with the problem of counselor

effectiveness criteria.


Counselor Effectiveness Ratings

The widespread use of global ratings of effectiveness as criteria

in counseling research has been previously mentioned. In their review

of the effectiveness literature, Walton and Sweeney (1969) point out

the broad base of acceptance rating scales have attained. Since the

original studies by Rogers and Dymond (1954) and the emergence of the

"Process Scale" conceptualized (1958) and published by Rogers (1959),

there have arisen a host of researchers and research instruments using

a ratings format.

In addition to the Carkhuff scales, already cited, similar rating

scales include the Self-Anchoring Scale of Counseling Effectiveness

(Kilpatrick & Cantril, 1960), the Communication Rating Scale (Brams,

1961), the Counselor Potential Scale (Dole, 1964), the Wisconsis Rela-

tionship Orientation Scale (Wasson, 1965), the Counselor Evaluation

Inventory (Linden, Stone, & Schertzer, 1965), the Counselor Evaluation

Form (Faulkenberry, 1968), the Counselor Practicum Evaluation Form









(Eberlein & Park, 1971), and the Counselor Evaluation Rating Scale

(Myrick & Kelly, 1971), which was one of two criterion measures employed

in this research study.

There have been a variety of other unnamed rating scales including

a sorting process (Embree, 1954), and free response type counselor

perception instrument (Arbuckle & Wicas, 1957), sociometric indices

(Gade, 1967; Stefflre, King, & Leafgren, 1962), a categorized rating

instrument (Keltz, 1966), a sensitivity scale (O'Hern & Arbuckle, 1964),

and a number of semantic differential rating scales (Bartlett & Thomp-

son, 1971; Redfering, 1973; Riese & Stoner, 1973; and Shapiro, 1968).

It can be said that rating scales have served a large number of

researchers as an accessible, simplified, and quantifiable approach to

the complex problem of finding a viable, valid effectiveness criterion.

In general, the above instruments were validated on the basis of accept-

able correlations with other measures of counseling effectiveness such

as rankings or "expert" judgments. At times validity has been inferred

from reliability studies. While such practices are defended by respected

researchers in the field (Combs & Soper, 1963; Russo, Kelz, & Hudson,

1964; Stefflre, King, & Leafgren, 1962), it should be noted that actual

client outcome criteria were not generally used in the validation of the

above instrument.

Although the use of rating scales has extended into many areas that

are not germaine to this study, it can be noted that considerable

research has been generated using counselor self ratings, peer ratings,

coached client ratings, independent judge ratings, and a wide spectrum

of combinations. Nevertheless, this study focused specifically upon

the use of supervisor and client ratings as the basis for effectiveness








criteria. Both supervisor and client ratings are explored in some depth

in the review which follows.


Supervisor Ratings

There appears to be a general consensus that supervisor ratings are

among the most valid measures of effective counseling (Patterson, 1964).

The assumption is made that highly experienced counselors who comprise

"expert" panels are the most qualified of all judges to identify differ-

ential levels of effective counseling. The logic of Cattell, penned in

1903, is still compelling today: "There is, however, no other criterion

for a man's work than the estimation in which it is held by those most

competent to judge" (p. 314). Clearly the determination of "those most

competent to judge" remains a debatable point, but Cattell's rationale

continues to be invoked to justify the use of supervisor ratings.

Supervisor ratings have been frequently employed as a criterion

measure of counselor effectiveness. Such ratings were used in studies

by Allen (1967), Bishop (1971), Brown and Cannaday (1969), Demos and

Zuwaylif (1966), Embree (1954), Friesen and Dunning (1973), Gade (1967),

Keltz (1966), McDougall and Reitan (1961), Watley (1967), Wicas and

Mahan (1966), and a host of others. For the most part interrater relia-

bilities in these studies have been within the acceptable range. We

note, for example, that in the study by Gade (1967) the interstaff rank-

ing achieved a reliability of .80 when computed by the coefficient of

concordance statistical formula. In their 1973 study Friesen and

Dunning found that supervisors had an interrater reliability of .94

on the Counselor Effectiveness Scale. This coefficient was higher than

the .92 achieved by lay people or the .91 reliability among practicum









students. Using the same instrument, Keltz (1966) found the interjudge

agreement on pooled ratings to range between .73 and .78 and concluded

that the instrument and the supervisor rating approach "constitute a

usable, realistic measure for assessing the counseling proficiency of

counselor-trainees" (p. 516). Likewise, when Falkenberry compared the

evaluation consistency of supervisor, peer, role-player, and self ratings

on the Counselor Evaluation Form, he concluded that for overall scores

and in specific items the supervisors were more reliable than any of the

other groups of raters.

A comprehensive review of published studies using the Counselor

Evaluation Rating Scale is presented in the next chapter. The CERS

(Myrick & Kelly, 1971) is one of the criterion measures employed in

this study and, therefore, is explored in depth under the section on

"Instrumentation."


Client Ratings

The use of client perceptions to evaluate counseling effectiveness

has been extensive, and the reactions of researchers overall tend to

qualify the approach as a valid criterion. The value of client ratings

is attested to by a substantial number of authors (Arbuckle, 1956; Form,

1955; Grigg, 1961; Grigg & Goodstein, 1957, 1959; Linden, Stone, &

Shertzer, 1965; Mueller, Gatsch, & Ralston, 1963; Patterson, 1958;

Pohlman, 1961; Pohlman & Robinson, 1960; Rosen, 1967; Severinsen, 1966;

Shoben, 1953; Stablein, 1962; Thompson & Miller, 1970).

An early cogent argument for the inclusion of client ratings in

effectiveness research was penned by Grigg and Goodstein (1957). Their

position was that:








Some appraisal of the client's reaction to the counselor and
to counseling should be obtained before we can say that we
have any comprehensive understanding of who makes a good
counselor a successful practitioner, among other things,
is one who elicits favorable reactions from the recipients
of his services. (Grigg & Goodstein, 1957, p. 31)

Indirect support for counselor ratings may be drawn from studies

which demonstrate significant positive correlations between client and

supervisor rankings of counselor effectiveness (Brown & Cannaday, 1969).

In several studies using the Counselor Evaluation Inventory (CEI) devised

by Linden, Stone and Shertzer (1965) high correlations between client

and supervisor ratings lend support to the hypothesis that clients and

supervisors tend to evaluate a similar set of criteria (Anderson &

Anderson, 1962; Correll, 1955; Poole, 1957). In research involving the

same instrument, Rickabaugh, Heaps, and Finley (1972) discovered a

significant relationship between clients' perceptions of counselor com-

fort and client academic grade improvement, another frequently used

measure of effective counseling outcome. Pfeifle (1971) also found that

clients rated counselors with practicum experience significantly higher

than counselors without practical experience when ranked on the CEI.

Client ratings have often been an integral component in the pre-

viously discussed research on the therapeutic conditions of empathy,

warmth, genuineness, respect, and concreteness offered by the counselor

(Truax & Carkhuff, 1967). One of the pioneers in using client percep-

tions to measure the therapeutic conditions was Barrett-Lennard (1962).

An extensive analysis of the characteristics and uses of his instrument

follows in the section of the next chapter on "Instrumentation." The

use of his instrument and of similar relationship inventories has yielded

positive results with client samples ranging from juvenile delinquents









(Truax, Wargo, Carkhuff, Turnell, & Glenn, 1966) to schizophrenics

(Rogers et al., 1967) to undergraduates seen at a university counseling

center (Kurtz & Grummon, 1972). The review of the broad scale applica-

tions of the Barrett-Lennard Relationship Inventory should be noted in

Chapter III.

The issue has sometimes been raised that subjectivity and bias tend

to invalidate client perceptions as a valid criterion of counseling

success (Patterson, 1958, 1959; Pohlman, 1961; Rosen, 1967). One

criticism has been that unrealistic expectations may lead to inevitable

dissatisfaction with a consequent blurring of objectivity (Patterson,

1958; Severinson, 1966). Some support for this position came from a

study by Pholman and Robinson (1960) who felt their research confirmed

the existence of definite counselor preferences among clients. The study

consisted of verbal presentation of 92 counselor behaviors to a group of

psychology students who rated the behaviors in terms of degree of

annoyance or liking. Mean ratings of 73 behaviors differed signifi-

cantly from the expected mean of neutrality. Furthermore, in a review

of the literature on client preferences, Rosen (1967) listed a substan-

tial number of other client preferences which may influence a client's

perception and subsequent rating of counselor effectiveness.

Another variable, client preferences for specific counseling

approaches or sets of counseling techniques, may also alter the client's

ratings. Greater satisfaction was reported by Barahal, Brammer, and

Shostrom (1950) for clients counseled by "client-centered counseling"

than for clients counseled by "traditional" techniques. One must be

cautious, however, in interpreting such findings because a client's

phenomenon influenced by a variety of confounding and interacting









variables. Among these variables are the client's actual problem

(Bordin, 1955), the personality of the client (Sonne & Goldman, 1957),

the degree of counselor activity (Forgy & Black, 1954), the identity of

the counselor (Grigg & Goodstein, 1957), and perhaps a gamut of other

possible variables. Patterson (1959) pointed out that one pernicious

danger in interpreting client ratings is the possible discrepancy be-

tween what a client wants and feels he needs and what is in the client's

best interest. It is conceivable that a client may rate a counselor

positively when in reality the counselor may be compounding the client's

problem. In fact, some studies have suggested that clients may not be

the most objective judges of the counselor's therapeutic conditions

(Burstein & Carkhuff, 1968; Truax, 1961).

In conclusion, the preponderance of research evidence supports the

use of client perceptions and ratings of counselor effectiveness.

Nevertheless, one must be mindful of the complex variables such as

preferences, expectations, and situational factors which may impinge

upon and qualify their meaning. Therefore, a certain degree of caution

is warranted in their use; namely, they should be employed in conjunc-

tion with multiple criterion measures such as supervisor ratings. In

calling for multiple criterion measures of counseling effectiveness,

Goodstein and Grigg (1959) concluded that "client satisfaction is one

important factor in any multifactor approach to the problem of effec-

tiveness in counseling" (p. 23).


Summary

Both supervisor and client ratings have been researched extensively

as effectiveness criteria. Supervisor ratings have enjoyed a high level






35

of acceptance and continue to be regarded as one of the most valid

criterion measures. Client ratings have yielded more mixed results, but

the weight of evidence supports their use. When supervisor and client

ratings are used in conjunction, as they have been in this study, there

is evidence that each strengthens the credibility of the other.















CHAPTER I II

METHODOLOGY


This study attempted to investigate the extent to which effective-

ness ratings of counselor trainees were related to preferred communica-

tional modes. Preferred verbal responses of counselor trainees were

examined in relationship to both client perceptions of counseling rela-

tionships and supervisory evaluations of counseling effectiveness.


Description of the Sample

The subjects involved in this study included University of Florida

counselor trainees, the trainees' clients, and supervisors of the

trainees. A more detailed description of these samples, including the

process of selection, follows.

I. Counselor Trainees. The counselors-in-training group in this study

were composed of graduate students in the Department of Counselor

Education at the University of Florida. These counselor trainees

were all currently enrolled in a practical, supervised counseling

experience, referred to as a practicum or an internship, through the

Department of Counselor Education. Prior to this supervised train-

ing period each trainee had completed an introductory course in the

helping process followed by a course which combined a survey of

counseling theories and approaches with a counseling laboratory.

The laboratory experience utilized role playing and video-taped

microcounseling (Ivey, 1971) with volunteer clients. Both of these

36










courses lasted for an academic quarter and preceded the first super-

vised practicum. Students pursuing a specialist degree must com-

plete three practice which typically amount to ten hours per week

and one internship quarter which is defined as a full-time, forty

hours per week counseling experience. For doctoral students the

total requirement is five practice and three quarters of intern-

ship. Counselor trainees in this study were at any point in the

completion of their degree programs from the first practice to the

last internship. Trainees have available actual counseling oppor-

tunities in a variety of academic and community settings. Counselor

trainees in this study were, for example, involved in a university

counseling center, a university mental health service, a community

mental health center, a drug treatment center, a delinquency treat-

ment program, a paraprofessional training program, a community

college, a public school, a hospital, or any of a number of similar

human service agencies either affiliated with the University of

Florida or in nearby communities.

Graduate counseling students who were involved exclusively with

elementary and middle school children were not included in this

study since children that young often have difficulty completing the

inventories used in this study. All other counselor trainees

involved in a practicum or internship during the target quarter,

and who received supervision through the department, were asked to

participate in the data collection process.

2. Supervisors. The supervisors in this study were counselor educators

or professional counselors who had themselves completed rigorous

graduate training programs in counseling at the university level.









All supervisors had extensive counseling experience in a variety of

settings with a wide range of clients. Most of these supervisors

had completed a counseling internship of a full academic year in

addition to other supervised practice or clerkships before entry

into full-orbed professional practice. Every supervisor was a

faculty member in the Department of Counselor Education or had been

approved by the department. Each supervisor held an earned doctorate

in an appropriate field such as Counselor Education or Psychology.

All of the supervisors had conducted weekly individual supervision

with the counselor trainee being evaluated for a period of not less

than six weeks.

3. Clients. The clients in this study were persons seen by the coun-

selor trainee in a helping relationship at the trainee's assigned

agency during the predetermined week of data collection. Each

client had to have been seen by the counselor trainee subject for

a minimum of two sessions of at least one half hour duration to be

included in the study.


Data Collection

Data collection for this study occurred during the seventh and

eighth weeks of the Spring, 1976, quarter at the University of Florida.

During that time frame each counselor trainee subject was evaluated by

an individual supervisor. Only counselor trainees who received super-

vision on the university campus or in the local community were included

in the sample. Also, evaluative ratings were made by each trainee's

individual supervisor only. Generally, the individual supervisor met

weekly with the trainee on a one-to-one basis for one hour per week.









The focus of supervision was to review the trainee's counseling efforts,

primarily via tape recordings and verbal report; to assess the trainee's

level of skill; and to guide the supervisee's development toward full

professional competence. The method used to evaluate the effectiveness

of the supervisee was the Counselor Evaluation Rating Scale (see Appen-

dix) which the supervisor returned to the researcher following comple-

tlion.

During the first week of data collection, the investigator arranged

with all relevant group supervisors to meet with their groups for the

purpose of administering the Helper Response Preference Inventory (see

Appendix) to each counselor trainee. At that time demographic data on

each trainee were also gathered by means of a questionnaire (see Appen-

dix). The first two demographic variables, sex and major field, were

included to obtain some descriptive data on the counselor trainee sample,

They were omitted from the statistical analysis since the data were not

continuous. Altogether the following six sets of demographic informa-

tion were requested:

I. Sex of the counselor Irain. ;

2. Undergraduate field of major;

3. Age of the counselor trainee;

4. Cumulative graduate grade poinl average at the end of the

previous quarters;

5. Number of graduate quarters completed at the end of the quarter

of data collection;

6. Number of quarters of supervised counseling practice completed

at the end of the data coll action quarter.

At each group supervision session the counselor trainees were








requested to administer the Barrett-Lennard Relationship Inventory to

one of their clients. The trainees were then informed about the adminis-

tration procedure. Specific verbal instructions to the counselor

trainees were as follows:

From among the clients you have seen in a counseling relationship

for a minimum of two thirty-minute sessions, select your most

typical client. Ask the selected client to please complete the

Barrett-Lennard Relationship Inventory based upon your counseling

contacts together. Make sure the client reads the cover letter

of explanation. Also, be certain that the client thoroughly under-

stands the rating procedure. Emphasize that he or she is the only

client you are asking so cooperation is extremely important.

Offer to stay with the client while the first page of the Rela-

tionship Inventory is completed in private, to answer any questions

that might arise. When you are convinced the client understands

the task and will follow through, stress the fact that the instru-

ment will be ready to mail as soon as completed and leave the client

to finish the ratings.

The client was then expected to comply with the verbal instructions

to evaluate and mail the instrument. Similar instructions were also

reinforced in the explanatory cover letter (see Appendix). The instru-

ment was stamped and pre-addressed so the client could personally mail

the Inventory to the researcher. The brief cover letter, in an effort

to maximize the probability of genuinely candid responses, explained the

procedure in part. It was emphasized that even though the ratings were

important, they would not be used to personally evaluate the individual

counselor trainee. Also stressed was the fact that since all information








is strictly for research purposes the ratings would be analyzed as group

data only.

In order to identify missing data, each Relationship Inventory

carried the name of the counselor trainee being evaluated. Five days

following the period of data collection, the researcher began calling

counselor trainees to identify and rectify the causes of unreturned

Relationship Inventories. Postcards were mailed to counselor trainees

who could not be reached at home to inform them of the missing Inventory

and request followup. Thirty days after the original week of data

collection it was assumed that all completed Inventories had been

received. Statistical analysis of the available data then proceeded.

Whenever counselor trainees were absent from their group supervi-

sion sessions, they were contacted by telephone to solicit their coopera-

tion. When the HRPI was administered and the Relationship Inventory

delivered to these trainees, the same instructions were given as in the

group sessions. Also, the follow-up of missing data was the same as

previously described for trainees instructed individually.


Limitations

Since a deliberate sample of counselor trainees was selected for

study rather than a random sample from the entire universe of counselor

trainees, the generalizability is limited to graduate counselor educa-

tion students at the University of Florida, their supervisors, and their

clients. Furthermore, the exclusion of elementary and middle school

counselors, although methodologically essential, further minimized the

external validity of the results obtained.

A second limitation of this study lies in the level of development









of the instruments employed as criteria of counseling effectiveness.

Neither the Barrett-Lennard Relationship Inventory nor the Counselor

Evaluation Rating Scale was validated directly upon client outcome

research. Therefore, the measurement of counseling effectiveness was

indirect, but, nevertheless, consistent with procedures generally

recognized as appropriate within the professional field of counseling.

One should not, however, construe the findings of this study to be

equivalent to validation based upon actual outcome measures of client

change. Likewise, the results of this study are specific to the instru-

ments used and should not be generalized to other comparable instruments

until further study demonstrates which global measures of counseling

effectiveness may be considered statistically interchangeable.


Instrumentation

The three instruments referred to in the presentation of the metho-

dology are each examined independently in the following section.


Helper Response Preference Inventory (HRPI)

The Helper Response Preference Inventory (HRPI) was originally

devised by Dr. James L. Lister in the early 1970's in an extension of

the work of Porter (1950) on classification of counseling responses.

The HRPI (see Appendix) consists of fifty items, each presenting a

statement by a helpee to which the respondent must choose one of two

optional helper statements as a response. The intent of the forced-

choice format was to pair twenty responses, of approximately equal

valence, from each of the five following categories: understanding,

supportive, interpretive, probing, and evaluative responses. Thus, the

one hundred total responses enabled each set of possible pair-choices









such as understanding-probing, or interpretive-evaluative to occur five

times each. The similar pairs were randomly distributed throughout the

test items. Theoretically, this method was believed to allow identifi-

cation of a counselor's preferred response style by quantifying the

over-chosen response categories.

Validity studies on the HRPI were conducted by Bradley (1973) in

an unpublished study of both counselor trainees and a paraprofessional

group of child care workers. Using the Counselor Evaluation Rating

Scale as a criterion, the regression studies for a sample of 65 counselor

trainees revealed that the HRPI scores and ratings on the CERS were

correlated .47 overall. The counseling scale of the CERS alone corre-

lated .48 with HRPI scores.

In the companion investigation, HRPI scores were correlated with a

global subjective rating of child care workers by their immediate super-

visors as effective, adequate, or relatively ineffective helpers. A

multiple correlation of .63 was yielded, but the understanding score

alone correlated .61 with supervisor rankings.

Test-retest reliability of the HRPI was found to be .83 for a

heterogeneous sample of 43 subjects after a span of four to six weeks

(Bradley, 1973).

Split-half reliabilities were computed by Lister (1972) for the

five original sub-scales as well as a composite scale based upon forty

consensus items agreed upon by a sample of eleven doctoral level coun-

selor educators on the faculty of the University of Florida. This con-

sensus scale was labeled the "expert" scale. The split-half reliabili-

ties for the a priori scales and the "expert" scale were generally

modest and are summarized in Table I for three separate sample groups.










TABLE I

SPLIT-HALF RELIABILITIES FOR THREE SAMPLES: GRADUATE LEVEL COUNSELOR
TRAINEES, PARAPROFESSIONAL COUNSELOR TRAINEES IN A COMMUNITY COLLEGE,
AND PARAPROFESSIONAL DISASTER RELIEF COUNSELOR TRAINEES



Santa Fe Community Counselor Education Wilkes-Barre Flood
College Human Service Graduate Students Paraprofessional
Aide Trainees Emergency Counselor
Trainees
(N=61) (N=72) (N=50)


Expert .62 .62 .49

Understanding .53 .52 .52

Supportive .41 .47 .42

Interpretive .17 .42 .34

Probing .18 .30 .52

Evaluative .13 .37 .29




Factor analysis of the HRPI was conducted in 1973 (Bradley) pooling

319 sets of responses in the sample. Since the items were forced

dichotomies, it was imperative to compute tetrachoric correlations. The

items were factor analyzed, and the principal axis solution rotated to

varimax criterion (Guertin & Bailey, 1970). Highest coefficients

served as communality estimates. Two items had to be deleted because

of low cell distributions. The remaining 48 item variables were factor

analyzed and again rotated to varimax criterion, but a clear factor

structure did not emerge until the principal axes factor matrices were

rotated in an oblique rotation procedure. Ten principal axes factors

were chosen because the latent roots tended to plateau beyond that point

and all ten factors chosen had latent roots greater than 1.0. The









oblique factor matrix was clear with ten factors accounting for 60.2%

of the total score variance and 87.6% of the common variance. Factors

were limited to items that had at least a .30 factor loading. Factor

scores were obtained then by weighting each item in proportion to its

factor loading as follows:

Factor Scores

Factor Loading Range Weight of the Item

.30 to .39 I

.40 to .49 2

.50 to .59 3

.60 to .69 4

.70 to .79 5

The relative independence of the oblique factor structure is

attested to by the low correlations between primary factors. All inter-

correlations are below .30 except between factors A and J, which is

0.38.

The 48 items, which were the variables factor analyzed, presented

alternative choices of responses perceived as being helpful. In each

item a value judgment was required by the respondent of which item to

choose and which to reject. Consequently, every response represented

a preference of one response over another one. The factors corres-

pondingly have a high end which represents the responses chosen and a

low end which represents the responses rejected. The factor labels,

which are believed to describe the common theme of the items loading in

each factor group, are listed below. Also presented are the number of

items loading the factor and the total weights for the factor. Factors

are listed in descending order according to the proportion of total

score variance accounted for by the factor.









TABLE 2

FACTOR LABELS AND FACTOR SCORES FOR THE HELPER
RESPONSE PREFERENCE INVENTORY


Preferred Response Rejected Response Number Total of
Factor Label Factor Label of Items Weights


A. Empathic Under- Probing 12 27
standing

B. Confrontation False Reassurance 9 21

C. Specificity Ambiguity 8 20

D. Validation of Analytical Interest 6 12
Experience

E. Clarification Superficiality 9 16

F. Focused Probing Premature Judgment 5 II

G. Tentative Formu- Avoidance of Emotion 4 7
lations

H. Requesting In- Giving Information 4 8
formation

I. Helpee's Intern- Helper's External 5 8
al Reference Reference

J. Frank Feedback Incongruence 5 10




Barrett-Lennard Relationship Inventory (BLRI)

The theoretical rationale which led to the construction of the

Barrett-Lennard Relationship Inventory (Barrett-Lennard, 1962) was

derived from the formulation by Carl Rogers (1957) of "necessary and

sufficient conditions" for beneficial client change. The Relationship

Inventory was intended to assess four aspects of any two-person rela-

tionship, particularly the counselor-client relationship. Those four

aspects were empathic understanding, level of regard, congruence, and









unconditionality of regard. The original instrument contained 85 items

but was soon shortened to 72 items. Most of the research on the Rela-

tionship Inventory before mid-1964 utilized the 72-item "interim revi-

sion" of the original investigative instrument. Since that time, the

64-item revision, or adaptations of this revision, has been employed

in most research using the Inventory. The author (1969) explains the

procedures used in revising the instrument to its current form:

The 64-item revision was based (a) on item-analysis results
from several samples of data obtained with earlier versions
of the instrument,(b) on other technical considerations,
such as achieving a numerical balance of positively and
negatively stated items in each of the four scales, (c) on
minor theoretical refinements, particularly in respect to
the unconditionality of regard scale and (d) on the con-
cern to alter or replace items of a relatively abstract
or difficult kind. The item analysis provided empirical
checks on what each previously used item was doing and
gave results that were used in conjunction with b), c)
and d) to refine the selection and preparation of items.
In general, the aim of the revision was to further im-
prove the sensitivity and versatility of the instrument
for measuring the defined variables in the context of any
significant interpersonal relationship. (Barrett-Lennard,
p. I, 1969)

Well over one hundred research studies have used the Barrett-

Lennard Relationship Inventory establishing the instrument as one

of the most widely used measures of the quality of human relation-

ship available today. In its current form, the Relationship Inven-

tory's 64 items include 16 for each of the four subtest dimensions,

which assess a person's perception of another person in a relation-

ship. Each statement is scored on a 6-point scale, from +3, "yes,

I strongly feel that is not true." There is no 0 response. The

items are general enough to assess practically any two-person rela-

tionship which the examiner may specify.

The published reliability studies on the earlier 72-item form








of the Relationship Inventory were acceptable and generally consistent

across investigations. Barrett-Lennard (1962) reported split-half

reliability coefficients for the various relationship dimensions within

a range of .82 to .93. In two separate studies, Snelbecker (1961,

1967) discovered a range of split-half reliabilities between .75 and

.94. The range was .83 to .95 for Hollenbeck (1965) and .82 to .91

for Hough (1965).

The test-retest reliability coefficients reported by Barrett-

Lennard (1962) after a four-week interval, ranged from .84 to .90. After

a six-months interim between administrations Hollenbeck (1965) found a

somewhat lower range of reliability, on the 72-item form, of from .61

to .81. In an investigation using the group form of the Relationship

Inventory, Berzon (1964) reported a reliability coefficient of .86 for

total scores after a four-week interval.

Refinement procedures have helped the reliability coefficients for

the 64-item revision of the Relationship Inventory to run as high or

higher than with previous versions of the instrument. Several test-

retest studies are summarized in Table 3.

The question of the validity of the Barrett-Lennard Relationship

Inventory is interwoven with the complexities of obtaining direct cri-

terion-based validity checks in the overall field of counseling.

Barrett-Lennard (1969) relates this difficulty to "the absence of alter-

native, established measures of theoretically equivalent dimensions of

perceived interpersonal response" (p. 4). Although a number of

researchers have been basically in agreement with the statement by Mills

and Zytowski (1967) that "the test has, at a minimum, face validity,"

(p. 194) a more persuasive justification for its use as a criterion

instrument is essential.








TABLE 3

TEST-RETEST RELIABILITY OF THE SUBTESTS OF THE
64-ITEM RELATIONSHIP INVENTORY
IN SEVERAL STUDIES


Mills and Mills and
Relationship Barrett-Lennard Barrett-Lennard Zytowski Zytowski
Dimensions (1969) (1969) (1967) (1967)
Form OS, Form MO, Form MO, Form OS,
N=40 N=38 N=79 N=79
2 to 6 weeks 12 days 3 weeks 3 weeks


Level of
Regard .88 .79 .86 .74

Empathic
Understanding .86 .91 .84 .90

Unconditiona-
lity of Regard .86 .86 .80 .80

Congruence .92 .85 .87 .88

Total Score .92 .89 --- --





It must be recognized that evidence supporting the validity of the

Barrett-Lennard Relationship Inventory is cumulative rather than con-

clusive. The bulk of evidence derives from research, much of which was

carefully designed, in which the associations between Relationship

Inventory measures and other variables were predicted on the basis of

logical and theoretical grounds. When these findings are positive, the

underlying theory gains credibility; and the argument that the dimen-

sions, purported to be measured, are indeed related to measurable rela-

tionship variables of fundamental importance in the consequences of

human interaction becomes more believable. The assumption, then, was

that a mounting body of findings which were consistent with the constructs









upon which the instrument was based established a defensible justifica-

tion for utilizing the Relationship Inventory as a criterion in this

study.

The basic theory underlying the Relationship Inventory (Rogers,

1957; Barrett-Lennard, 1962) has been substantially supported by the

vast range of studies using the Truax and Carkhuff rating scales to

measure levels of therapeutic conditions within the context of relation-

ship (Truax & Carkhuff, 1964, 1967; Carkhuff & Truax, 1965a, 1965b;

Berenson, Carkhuff, & Myrus, 1965).

More pointed research has involved the instrument per se. Success-

ful use of the Relationship Inventory to appraise counselor-client rela-

fionships has been reported by Van der Veer (1961), Barrett-Lennard

(1962), Cahoon (1962), and Gross and Deridder (1966). Other intimate

relationships measured with significant results include marital partners

(Thornton, 1960), student-parent (Hollenbeck, 1965), supervisor-teacher

(Blumberg, 1968), and student-teacher (Mason and Blumberg, 1969).

Barrington (1961) related scores on the Barrett-Lennard Relation-

ship Inventory to such specific counselor behaviors as a decrease in

the length of counselor responses from the first to the second inter-

view, percentage of emotional words used by the counselor, and simi-

larity of the counselor's rate of speech to the client's rate. The most

predictive sub-scale was the empathic understanding dimension, but

perceived counselor congruence was highly related to more integrated

verbalizing by the counselor in the second hour of counseling.

Using a Q-sort technique, Emmerling (1961) studied the relation-

ship of "openness" in relatively effective and ineffective teachers to

how their students perceived them in relationship. It was hypothesized








that "open" teachers would see themselves as responsible for diffi-

culties and remedial action in their classroom work. The rationale was

that such teachers would tend to be more genuine, sensitive, and

accepting than teachers who, under similar conditions, see problems as

unrelated to their own characteristics and externally imposed. The

finding that students of the more "open" teachers did perceive them more

positively on each of the Relationship Inventory dimensions may be

interpreted as implying that the measures discriminated differences

consistent with theoretical predictions.

Cahoon (1962), in another well-known investigation, found signifi-

cant relationships between the process experiential levels and open-

mindedness, on one hand, and the client-perceived quality of the coun-

seling relationships on the other. For his sample of practicum-level

counselor trainees he employed the Process Scale, the Dogmatism Scale,

and the Barrett-Lennard Relationship Inventory.

The uses of the Relationship Inventory have been extensive.

Donnan, Harlan, and Thompson (1969) used the original 85-item form to

significantly correlate four personality factors on the 16 PF with

three relationship variables for a sample of 22 counselors who were

evaluated by 880 prospective college freshmen following counseling. In

another study, Brown and Calls (1968) measured the effect upon the

client's perception of the relationship with his or her group counselor,

depending upon whether the group was required or self-initiated. Lanning

(1971) used the Relationship Inventory to evaluate whether practicum

students perceive the supervisory relationship the same in group and

individual supervision. A still different application of the Inventory

was made by Armstrong (1969) who examined the difference between students








in counseling and students not in counseling with regard to quality of

their relationships with intimate friends and the tendency to use

intimate friends as therapeutic agents. In a very recent study Loesch

(1975) used the Relationship Inventory to assess whether a client's

nonverbalized feelings and the counselors' perception of them during

the interview were related to subsequent counselor and client percep-

tions of the counseling relationship.

The appropriateness of using the separate relationship dimensions,

purported to be measured by the Relationship Inventory, rather than the

more global total score, has been the subject of some controversy.

Initially Barrett-Lennard (1962), in asserting that separate relation-

ship aspects were indeed being measured, stated that the scales were

not measuring a general factor, such as the client's overall satisfac-

tion or dissatisfaction with the therapy relationship. However, Mills

and Zytowski (1967), on the basis of a principal-components analysis

of the intercorrelations between the subtests for several samples using

a variety of forms of the Inventory, concluded differently. They found

that a general component accounted for about two-thirds of the variance.

Walker and Little (1969) challenged the findings of Mills and

Zytowski since their conclusions were based upon an unrotated principal-

components analysis. Walker and Little then combined the Inventory

scores of 150 university undergraduates and conducted a component

analysis of the 64 items. They found three factors which they labeled

"non-evaluative acceptance," "psychological insight," and "likability."

At about the same time Tosi, Frumkin, and Wilson (1968) did inter-

correlational studies on the subtests of the Relationship Inventory

scores of 69 clients at the Kent State University Guidance Laboratory.








They concluded, from the correlation range between subscales of from

.31 to .58, that their "data appear to support Barrett-Lennard's con-

clusion that the four subscales are relatively independent" (p. 642).

In 1974 Lanning and Lemons administered the OS form of the Rela-

tionship Inventory on three separate occasions to compile data for a

principal-components analysis and an orthogonal rotation of the factor

matrix by the varimax procedure. They found two nearly orthogonal

factors with Level of Regard, Empathy, and Congruence loading on a

component that accounts for from 62 to 72 percent of the total variance.

Unconditionality of Regard loads predominantly on the other factor which

accounts for 16 to 21 percent of the total variance. The authors felt

their study provided some support for the earlier findings of Mills and

Zytowski. Lanning and Lemons concluded:

The major component is probably a measure of the overall
satisfaction with the relationship, contrary to earlier claims
made about the separateness of the subtests. This indicates
that there may be one predominant characteristic across all
of the relationships measured rather than four separate ones.
(1974, p. 230)

Since there appeared to be somewhat inconclusive empirical support

for or against a general relationship satisfaction factor, this study

employed the total score of the Barrett-Lennard Relationship Inventory

and each sub-score as criterion measures based upon client ratings of

perceived counseling effectiveness. Both the male and female forms of

the OS-64 revision (see Appendix) were utilized in this study.


Counselor Evaluation Rating Scale (CERS)

The Counselor Evaluation Rating Scale (see Appendix) which was

used in this study as a criterion measure of supervisor rating, com-

prises 27 items that enable a supervisor to rate a counselor trainee's








performance in a supervised practicum (Myrick & Kelly, 1971). The items

are designed to assess the counselor trainee's understanding of a

counseling rationale, counseling practice with counselees, and explora-

tion of self and counseling relationships with a supervisor. A 7-point

(+3 to -3) Li, kert-type scale is employed to mark each item. The CERS

yields three scores: a counseling subscale (C) is an appraisal of the

counselor-trainee's individual counseling behavior; a supervision sub-

scale (S) is an assessment of the person's involvement and progress in

the supervisory process; and a total score (T) of these two subscales

plus one additional item: "Can be recommended for a counseling posi-

tion without reservation." The resulting total score ranges between 27

and 189 when the negative scores are eliminated. The counseling scale

and the supervision scale each yield scores between 13 and 91.

In their original article (1971) the authors, Myrick and Kelly,

reported a reliability study involving 45 student counselors and their

supervisors from the University of Florida. Individual supervisees,

who met with their assigned faculty supervisor for a minimum of one

hour per week for twelve weeks, were rated by these supervisors on the

CERS at the end of a fall quarter practicum. Utilizing the Spearman-

Brown correction of the split-half reliability procedure, the investi-

gators found a coefficient of .95. When the 13 counseling items were

compared with the 13 supervisor items, a correlation coefficient of .86

was produced.

The stability of the CERS over a period of time was determined by

a test-retest reliability procedure. After a minimum of a four week

lapse in time since collecting the CERS ratings, the supervisors were

again requested to rate the same student counselors with whom they had








been involved during the fall quarter. Data analysis yielded a .94

product moment reliability coefficient.

Myrick and Kelly (1971), in their presentation of the instrument,

reported they had failed to find relationships between either Graduate

Record Examination scores or undergraduate grade point averages and CERS

ratings. The correlation coefficients were respectively -.03 and -.25.

Wittmer and Lister (1971) also correlated Graduate Record Examina-

tion scores and CERS ratings for 53 counselors in training whose CERS

ratings were compiled following the first quarter of practicum. The

GRE had been taken prior to admission. The correlation of the GRE

verbal score and CERS ratings was -.08. When the GRE quantitative score

was correlated with the CERS, the coefficient yielded was .12. Neither

correlation was significant at the .05 level. However, this same study

also compared results of each counselor trainee's 16 PF profile, taken

the quarter before the first practicum, with his CERS ratings. The

authors employed the index of predicted counseling effectiveness which

is a regression equation based upon studies of the 16 PF at the Insti-

tute for Personality and Ability Testing (1963). The Pearson product-

moment correlation coefficient produced was .41, which was significant

beyond the .01 level.

In another published article using the CERS as a criterion of

counseling effectiveness, Myrick, Kelly, and Wittmer (1972) found that

scores correlated significantly with certain factors of the 16 PF. The

investigators defined 20 effective counselor trainees with high CERS

scores and 20 ineffective counselor trainees with low CERS scores. The

15 trainees with scores near the mean of the CERS were eliminated. An

examination of the 16 PF scores of the two comparison groups yielded








eight personality factors with score differences between the groups

which were statistically significant.

In a challenge-of the previous research, Jones (1974) questioned

the appropriateness of using an instrument designed to help supervisors

evaluate practicum students' behavior as a criterion of counseling

effectiveness. He argued, "To be valid as a criterion measure of coun-

selor effectiveness, an instrument such as the CERS must be related to

counseling outcome or some other aspect of job performance" (p. 113).

Following his line of reasoning, Jones attempted to investigate the

relationships between the CERS and empathy, genuineness, and respect

using the Carkhuff Scales since these variables "have been established

as being significantly related to counseling outcome" (p. 113). He

found a significant positive relationship only between the level of

respect offered in the counseling situation and supervisor ratings on

the S (supervision) scale of the CERS. He concluded:

there appears to be no relationship between levels
of empathy, respect, and genuineness and the other CERS
scales, although the positive correlations between respect
and the CERS scales of C (counseling) and T (total) may
be noteworthy. (p. 115)

Forster and Hamburg (1976) in a recent study of 30 full-time coun-

selor trainees failed to find a significant relationship between CERS

ratings and 16 PF scores.

A very recent study by Loesch and Rucker (1976), which has been

submitted for publication, has provided some support for the original

CERS Scales. In a second order factor analysis of more than four

hundred sets of ratings they obtained an oblique solution which closely

approximates the original classification of items into supervisory and

counseling scales.










In the report on validity studies using the Helper Response Prefer-

ence Inventory, it was pointed out that for a sample of 65 counselor

trainees there was a .47 correlation between HRPI scores and overall I

CERS scores. The correlation was .48 for the counseling scale of the

CERS when related to the HRPI scores separately.

Borman and Ramirez (1975) compared the CERS self ratings of 25

master's degree counseling students in practicum with their ratings by

their supervisor and practicum assistants who were doctoral students.

Their findings revealed that there were significant differences found on

9 of the 27 items. However, most of the differences were attributed to

the students rating themselves higher than supervisors or assistants

and, specifically, differences in ratings between students and practicum

assistants. On only one item, dealing with the degree of relaxation and

comfort in the supervisory sessions, was there a statistically signifi-

cant difference between the supervisor's ratings and the counselors'

self-ratings.


Analysis of Data

The data collection process, previously described, yielded four

basic groupings of data: demographic variables, Helper Response Pre-

ference Inventory scores, Barrett-Lennard Relationship Inventory scores,

and Counselor Evaluation Rating Scale scores. Using the two criterion

instruments eight dependent criterion variables were isolated: the

Relationship Inventory total score plus the four subscores, and the

counseling, supervisory, and total scales from the Counselor Evaluation

Rating Scale. By means of the SPSS library computer program for stepwise

regression (1973), the ten factor scores of the Helper Response









Preference Inventory and the four demographic factors were predictor

variables regressed upon each of the dependent criterion variables. The

stepwise solution program (Kerlinger, 1973) allows at each step the

variable that makes-the greatest increment to the multiple coefficient

2
of determination, R to enter the computational process and calculates

a regression equation for that step. Also at each step of the analysis,

several calculations are performed and printed. These include the mul-

tiple correlation coefficient, the regression equation, partial correla-

tions, and the regression and residual sums of squares. One advantage

of this stepwise program is that the investigator can specify the level

of F ratio for entering a variable into the equation and the level of

F for removing a variable from the equation. For this study an alpha

level of .05 was considered the level of significance necessary to fail

to reject the hypotheses.

Utilizing the stepwise regression procedure made it possible to

ascertain which combination of independent variables, HRPI scores,

demographic data or combinations thereof, was the most efficacious pre-

dictor of the dependent criterion variables, the Relationship Inventory

or the Counselor Evaluation Rating Scale. The regression equation

computed specified the precise beta weights for the independent variables

which, taken in combination, yielded the greatest reduction in the error

sums of squares when predicting either of the criterion variables.

This method of analysis also permitted an examination of the R2's

and F ratios produced by each variable's increase in the overall pre-

diction process. By observing the increment to prediction for each

variable, the relative efficacy of different variables in the regression

equation became apparent. In effect, each hypothesis was tested for the







optimum combination of predictor variables as well as for any separate

predictor variable which was, by itself, a significant predictor of

either of the dependent variables.

While not central to the purposes of this study, canonical corre-

lations between all ten HRPI variables and the four Barrett-Lennard

Relationship Inventory subscore variables were computed to establish

the maximum correlation possible between the two sets of variables and

to estimate the variance shared by the two composites of variables. The

same procedure was used with the three Counselor Evaluation Rating

Scale scores and the ten HRPI variables. Finally, canonical correla-

tions were computed for the four Barrett-Lennard Relationship subscore

variables and the three Counselor Evaluation Rating Scale score varia-

bles. An intercorrelation matrix of all the variables in this study

is presented on page 67. It was anticipated that an intercorrelation

matrix could help clarify the results and could possibly explain the

presence or absence of statistically significant relationships among the

variables under scrutiny.
















CHAPTER IV

RESULTS OF THE STUDY

Introduction


The research procedures for this study were enacted according to

the experimental design described in Chapter III. Analysis of the data

was conducted via the computer facilities of the Northeast Regional Data

Center under the auspices of the Florida State University System. Spe-

cifically, the stepwise regression subprogram and the canonical corre-

lation subprogram of the Statistical Package for the Social Sciences

developed by Nie et al. (1975) were the analytical procedures utilized.

The results of these statistical operations are presented below in a

tabular format. This chapter consists of: (I) a descriptive profile

of the counselor trainee sample, including demographic data on the

subjects, and the scores of all subjects on the two criterion instru-

ments as well as the predictor instrument; and (2) the statistical tests

of the two major hypotheses and eight minor hypotheses subsumed under

the two major ones, presented in sequential order.


Descriptive Data of the Sample


The Counselor Trainees

There were 67 counselor trainees who appeared to meet the criteria

for inclusion: namely, having clients on a continuing basis who

possessed the literacy skills requisite to valid completion of the

60









Relationship Inventory. Of those 67 counselor trainees, two chose not

to participate for personal reasons, one was unavailable during the data

collection period, and three failed to return even the demographic data

sheet and HRPI answer sheet after agreeing to do so, despite follow-

up attempts. At least partial data were compiled on 61 counselor

trainees. Supervisors' ratings were obtained on all 61 subjects in

addition to the demographic data and the HRPI answer form. However,

the client rating on the Relationship Inventory was returned on only

46 counselor trainees. Therefore, the net sample for this study was

46 counselor trainees.

Descriptively, the counselor trainee sample was composed of 27

males and 19 females. The age range was 22 to 51 years, and the mean

age was 27.6. The graduate grade point average ranged from 3.20 to 4.00

with a mean of 3.77. The counselor trainees in the study had completed

from 3 to 17 graduate quarters and I to 8 quarters of supervised prac-

tice with respective means of 7.6 and 4.1. A breakdown of demographic

data on the counselor trainees is presented in Table 4.

The counselor trainees came from a wide variety of undergraduate

major backgrounds. Although the majority of trainees (27) had chosen

psychology as an undergraduate major, eight additional fields were

represented. There were five trainees each from education and language

undergraduate programs. Likewise sociology, religion, and the physical

sciences were represented by two counselor trainees each. There was

one counselor trainee each from political science, art history, and

business. The overall impression is that the sample represents a

diversity of backgrounds, interests, and personalities.









TABLE 4

DEMOGRAPHIC DATA ON THE 46 PARTICIPATING
COUNSELOR TRAINEE SUBJECTS


Variable

Age


Mean
Standard deviation
Median
Range


27.57
5.24
26.00
22-51


Sex
Ma I e 27
Female 19
Grade Point Average
Mean 3.7
Standard deviation .24
Range 3.20-
Number of Graduate Quarters Completed
Mean 7.61
Standard deviation 3.2l
Range 3-17
Number of Quarters of Supervised Coun
Mean 4.1
Standard deviation 2.24
Range 1-8
Undergraduate Major Field
Psychology 27
Education 5
Language 5
Sociology 2
Physical Sciences 2
Religion 2
Political Science I
Art History I
Business I


(58.7%)
(41.3%)


7
0
-4.00


3
6


selling Completed
0
6




(58.7%)
(10.9%)
(10.9%)
( 4.3%)
( 4.3%)
(4.3%)
( 2.2%)
(2.2%)
( 2.2%)









Supervisors

The supervisors in this study numbered 27. There were 20 males and

7 females. With the exception of one psychiatrist all of the super-

visors held an earned doctorate in counseling, guidance, student personnel

work, or some allied branch of psychology. Two supervisors gave

individual supervision to four trainees, and therefore each rated four

counselor trainees. Three supervisors completed ratings on the three

trainees they supervised individually, and seven supervisors rated two

trainees. The remaining fifteen supervisors each contributed a single

set of ratings on one counselor trainee. Only two female supervisors

completed more than one trainee's ratings, and in both cases the number

of ratings was two.


Clients

To intentionally preserve a sense of anonymity there was no attempt

to gather descriptive information on the client sample. While such knowl-

edge would of course be interesting, there would be no way to compare

the clients who responded with those who for one reason or another never

returned the Relationship Inventory. Consequently, client data would be

of limited interpretive value.


Client Evaluation Scores on the Counselor Trainees

Converted scores from the evaluative raw data are presented in

composite form in the following two tables. The client ratings on the

Barrett-Lennard Relationship Inventory are collapsed into means,

standard deviations, and ranges in Table 5. The scores are character-

ized by broad ranges with 46 to 47 points of spread among the four sub

scales. The total score range is 182 points.








TABLE 5

MEANS, STANDARD DEVIATIONS, AND RANGES OF SCORES ON THE
BARRETT-LEtJrJARD RELATIONSHIP INVENTORY (N=46)



Relationship
Standard
Dimensions Means Deviations Range

Level of Regard 35.83 10.27 2-48

Empathic Understanding 30.54 12.84 1-48

Unconditionality of Regard 21.04 12.02 1-48

Congruence 33.52 11.50 1-48

Total Score 118.35 42.91 10-192





Of interest is the fact that the means of the sub scales all fall

in the low to mid thirties except for the Unconditionality of Regard

score which dips to 21.04. Nevertheless, there is no appreciable

difference in the standard deviations of all four subscores which fall

in the range of 10.27 to 12.84. The Total score had a mean of 118.35

and a standard deviation of 42.91.


Supervisor Evaluation Scores on the Counselor Trainees


The individual supervisor ratings on the Counselor Evaluation

Rating Scale are presented in Table 6. The summary shows the means,

standard deviations, and ranges of the scores for all 46 counselor

trainee subjects. Ranges on both the Counseling and Supervision sub-

scales are quite broad, 33 to 91 and 38 to 91, respectively. The Total

score reflects a 91 to 189 point range.

The means of the two subscales, 81.33 for Counseling and 84.22 for








TABLE 6

MEANS, STANDARD DEVIATIONS, AND RANGES OF
SCORES ON THE COUNSELOR EVALUATION
RATING SCALE (N=46)


CERS Standard
Scales Means Deviations Range

Counseling 81.33 10.35 33-91

Supervision 84.22 9.98 38-91

Total 171.54 20.05 91-189





Supervision, are comparable. Both mean scores fall slightly less than

one standard deviation below the maximum possible score, a phenomenon

observed as well for the 171.54 mean of the Total scores.


Scores of the Counselor Trainees on the Predictor Instrument

The means, standard deviations, and ranges of scores on the ten

factor scales of the Helper Response Preference Inventory are presented

in Table 7. It should be kept in mind that each factor represents a

chosen direction of responses and a rejected direction of responses.

Higher scores reflect a preference for the response group named first in

the parenthetical pair, and lower scores indicate that the response

group placed second in the parenthesis was preferred. All of the means

fall in the upper half of possible score range for the factor except

for Factor D (Analytical Interest Validation) where the preference

favored slightly the validation side of the dichotomous scale. For

four factors the entire score range observed fell in or very close to









TABLE 7

MEANS, STANDARD DEVIATIONS, AND RANGES OF SCORES ON THE
HELPER RESPONSE PREFERENCE INVENTORY (N=46)


HRP I
Standard
Factors Means Deviations Range

A. (Empathic Understanding-
Probing) 24.22 4.11 !7-27

B. (Confrontation-False Reas-
surance) 16.26 3.21 8-21

C. (Specificity-Ambiguity) 18.11 2.68 10-20

D. (Analytical Interest-Vali-
dation) 4.13 2.60 0-10

E. (Clarification-Superficiality) 13.57 2.35 8-16

F. (Focused Probing-Premature
Judgements) 8.02 2.68 I-11

G. (Tentative Formulations-
Avoidance of Emotion) 5.76 1.75 2-7

H. (Requesting-Giving Information) 4.61 1.48 3-8

I. (Internal-External Frame of
Reference) 5.17 2.03 0-8

J. (Frank Feedback-Incongruence) 8.52 1.55 5-10





the upper half of possible scores. Three factors (D, F, and I) yielded

scores that covered almost the entire possible range. Factors B, G,

and H produced scores somewhere between these two extreme groups. No

other discernible pattern appeared to emerge from this set of data.


Relationships Among the Three Instruments

Table 8 presents the intercorrelation matrix for the two criterion












ko tm 00 rnt 0
(N 0 CM t4 CN 0
00000-










000-o'T0-
00C C00 -C


r- r- o

00-


0-

co
Z
0

<

LU
-r

00
cr z
< <
Z
Z
LUj U >-
-I -I w
I <0
-- 0 F-


wow
(-LO z



Z Z
Of C)
coc
u--w
< UJ
LU --. n
Z





.Zwu
I-J(2 L
0 w
LL LU
0 F- LL.
< LLU
LU I 0-


-J <
o > LU
< LUi CO




-Z
Qr IX 0
< 0 a-
O w -







HZ
< Z: w:
0 LUL
0 -J
LL LU LUI
m
I F-
'r- -
X I-

F- F--
0
< Q
2 0
F-

0 w



LU
cr--


IZ
0
C)
LUJ
t-
z


to
c
0

L -I--
+--
_0 -rC -0
L +- C:
(0 (a 0
m) 0. 0
() E c
rQ" LU =


0) (D
o +-
-3
L_ --


0 0
H) F-


CT


(n
oS


0
0


c
0

U)
>
L

:3
V) <0 c C)


- -- U ) cU)--- - - -
01 of Of o W ofa-. cLoCLcL cLL n 0_ a_ a_ 0_
-j -l -i -l -IJ w w w af cr-Of Of Qo r Q :Lry
coa commuC00CTaaTTa:m


(f) -
cr (o
LI+-
00
I--


Kt~ Nr- 0
0% r~- o












cooo

0 -







0
0
(D



CD



















%o 0 in rn '41 (N
0 0 0 0 NM
0000000
I I






0--N-- 0
0-- -O n -
0 0 N C
0000000
I I


rn 'rc 00
Co r~i o CMi o
cool-
0000--
II I


- 00 ON
CN CM -
000


- NM < m 01, K) IT
000---- 0O

00000000-
II


0 n r- NCM \CO O Pe)O
- CM o Q (D
co o o o os 6 o o











-("JO -o 00











Cd CM n m n CM o
-0 N CO 0
00000000-S










000000--
I III



-i-


0 )
-4-- c
>-.- )
4- C
-O -C -o S
- L +- C L -
(m mD U 0 O) co
o- CD- U C -
0 0 0 0 0
r- LU C) U--

Q Q Cr Of G Of
r"rv r r- rU"
- I -I j_1 -. J Ij
rCO CD 03 0 CO0c


C
0) 0
-U)
a) >
(n L-
c o
(flL

0C)
U) -- -- ----- -

QY: L CL0- CCL 0- m-C m m m mQ_
Sw I of O O cr Of O
0XIII II III

















0000
01\ r~- K)
- IZJ- -






0000




0000
0000
I I I I 6


(\---0-0-00000-0

0000000000000-
I I








Co OoO -O-N- 0
000-0 C--- CN0- 0
000000000000-
1


0 %4- "zr r o n (,J ON k)minoU\0 ) ^t 0
sNO-0O-O --OOO 000
000000000000000-
I I I I I I I I I I I I I








- (----0 N0a-),0
00--O0000--n-N00
00000000000000-
I I I I I I I


0---
0000
Illi


f- -r 0 00-r-- 0 -rn0
-ooo--O0--o
000000000 -
I I


in "rsim- Cl\r- m 0 ovooc No
0-000 ( n N C 00
000000000000-
I I




t-


0
t-

4- C
-- 0)

-0o EC 0 0
-- L 4- C L -
(0 (0 (x 0 C ) Of )
4- 0" Or U C 4-


--I -I 1 i L
-I J -i -IJ I LL
m mD co mn o


C
cJ)
Oo0


0 >

c )
='1 Q_

0 -3
OU)< 03 I UJL-OI--
V) - - -
of Ena.- Qo- Q L m -C m cL m -C
w w nluQx x x Of Ofr Orar
wwQIcIIIIIII








instruments, the Barrett-Lennard Relationship Inventory (BLRI) and the

Counselor Evaluation Rating Scale (CERS), and for the predictor instru-

ment, the Helper Response Preference Inventory (HRPI). There are 204

positive correlations and 120 negative correlations.

Significant in this matrix are the consistently negative correla-

tions between BLRI scores and CERS scores. The mean correlation for all

correlations between the two criterion instruments is -.21. All correla-

tions are negative except for the correlation between the BLRI Level of

Regard score and the CERS Supervision score which is a positive, though

ngligible, .02.

Canonical correlations were computed between subscale scores of the

BLRI and subscale scores of the CERS. Similar operations were performed

relating the HRPI variables to each set of the criterion instrument

subscales. In all cases the canonical correlations failed to reach the

.05 level of significance.

The relationships of each of the ten HRPI factors to the combined

scores of the BLRI and also to the combined scores of the CERS are

presented in Table 9. The statistic was derived by averaging the

correlations of the predictor HRPI factor with all subscales and total

score of each criterion instrument. The first column reveals the mean

correlation of the specified HRPI factor with the five subscores of the

BLRI. The second column shows the mean correlation of the specified

HRPI factor with the three subscores of the CERS. Notable are the seven

correlations, either positive or negative, above .15. Using this global

measure the .20 correlation of Factor E with the CERS scores and the .21

correlation of Factor C with the CERS scores seem worthy of more examina-

tion. Likewise the correlations of Factors H and J with the BLRI scores











TABLE 9

MEAN CORRELATION OF EACH HRPI FACTOR
WITH ALL OF THE SCORES OF THE
TWO CRITERION INSTRUMENTS


Mean of Correlation with All Scores
HRPI Factor BLRI CERS


Factor A (Empathic Understanding -
Probing) .16 .07

Factor B (Confrontation -
False Reassurance) .01 .17

Factor C (Specificity Ambiguity) .01 .21

Factor D (Analytical Interest -
Validation) .05 -.13

Factor E (Clarification -
Superficiality) .06 .26

Factor F (Focused Probing -
Premature Judgments) -.16 .04

Factor G (Tentative Formulations -
Avoidance of Emotions) -.01 .00

Factor H (Requesting Giving
Information) -.18 -.04

Factor I (Internal External Frame
of Reference) -.03 .08

Factor J (Frank Feedback -
Incongruence) .18 .12




on the order of -.18 and .18 respectively merit consideration despite

the lack of precision afforded by this crude statistical treatment.










Testing the Hypotheses


Test of the First Major Hypothesis

The forward stepwise multiple regression technique was applied to

all scores of the two criterion instruments, the BLRI and the CERS,

using the ten factor scores of the HRPI as predictor variables. Four

demographic variables (age, grade point average, number of graduate

quarters completed, and number of quarters of supervised counseling

practice completed) were also entered into the computations as possible

predictors.

The first hypothesis postulated that higher client ratings could

be predicted from counselor trainee response preferences as measured

by the HRPI, the predictor instrument. Table 10 presents the predictors

that met the .05 level of significance required for inclusion in the

regression equation for each of the five BLRI scores. The strongest

predictor overall was graduate grade point average (GPA). In fact GPA

was the best predictor of the Total score in addition to three of the

subscores. GPA was the only variable which significantly predicted the

Congruence subscale. The GPA accounted for 8% of the score variance in

the Level of Regard, the Empathy Level, and the Congruence scores. The

variance accounted for by the GPA was 11% of the Unconditionality of

Regard score and rose to 13% of the Total score. No other demographic

variable was a significant predictor of any of the BLRI scores.

HRPI scores were significant predictors on four of the five BLRI

scores. Factor F (Focused Probing Premature Judgments) accounted for

5% of the variance in both the Total score and the Empathy Level score.

In both cases Factor F explained a lesser amount of variance than the











TABLE 10

BARRETT-LErJA.F:D RELATIONSHIP INVENTORY PREDICTORS


2Criterion Predictors Multiple R R2 df
Criterion Predictors Multiple R R df F


Total Score

Graduate Grade Point Average

Factor F (Focused Probing -
Premature Judgments)

Level of Regard

Factor J (Frank Feedback -
Incongruence)

Graduate Grade Point Average

Factor A (Empathic Under-
standing Probing)

Factor I (Internal External
Frame of Reference)

Empathy Level

Graduate Grade Point Average

Factor F (Focused Probing -
Premature Judgments)

Unconditionality of Regard

Graduate Grade Point Average

Factor J (Frank Feedback -
Incongruence)

Congruence

Graduate Grade Point Average


0.35


0.42




0.41

0.50


0.53


0.57



0.28


0.36



0.33


0.41



0.30


0.13 1,44 6.34*


0.18 2,43 4.61*


0.17

0.25


1,44

2,43


8.70*

7.06*


0.29 3,42 5.58*


0.32 4,41 4.83*



0.08 1,44 3.72*


0.13 2,43 3.28*


0. II


1,44 5.44*


0.17 2,43 4.42*



0.09 1,44 4.27*


*p<.05









primary predictor and only other predictor, the GPA. Also in each case

Factor F was a negative predictor.

The single strongest predictor in this study related to the Level

of Regard score. For that score Factor J (Frank Feedback Incongruence)

accounted for 17% of the score variance. GPA added another 8% to the

prediction process. Factor A (Empathic Understanding Probing) and

Factor I (Internal External Frame of Reference) each explained 4%

of the variance. As a result the multiple correlation rose to .57, the

highest in this research, and collectively the four predictors accounted

for 32% of the Level of Regard score variance.

In addition to the 11% of variance attributed to the GPA in the

Unconditionality of Regard score Factor J (Frank Feedback Incongruence)

again entered the equation in the amount of 6% of variance accountability.

However, it should be noted that in this case, the variable which was

the strongest positive predictor of Level of Regard became a negative

predictor of Unconditionality of Regard.

Although the results are generally modest, ostensibly somewhat

contradictory, and involve only four of the ten HRPI factors, and only

four of the BLRI scores, there does appear to be a significant, obser-

vable trend between client ratings and response preference scores on the

HRPI. Therefore, with due regard to the need for further clarifica-

tion, the directional hypothesis HOI is accepted for the purposes of

this study.


Test of the Second Maior Hypothesis

Even though the four demographic variables were also used in the

regression of the second criterion, supervisor ratings on the CERS, no








demographic variable, including GPA, showed up in the set of signifi-

cant predictors found in Table II. It will be noted that for the CERS

Supervision score no predictor reached the necessary .05 level of

significance.


TABLE II

COUNSELOR EVALUATION RATING SCALE PREDICTORS
(p<.05)


Criterion/Predictors Multiple R R2 df F


Total Score

Factor E (Clarification -
Superficiality) 0.2721 0.0740 1,44 3.52*

Counseling Score

Factor E (Clarification -
Superficiality) 0.3208 0.1029 1,44 5.05*

Factor D (Analytical Interest -
Validation) 0.3836 0.1472 2,43 3.71*

Supervision Score

No Predictors


demographic variable, including GPA, showed up in the set of significant

predictors found in Table II. It will be noted that for the CERS Super-

vision score no predictor reached the necessary .05 level of signifi-

cance.

The CERS Total score was predicted by only Factor E (Clarification -










Superficiality) which accounted for 7.4% of the variance. With only

one predictor the simple correlation and the multiple correlation were

an identical .27.

There were two factors predicting into the CERS Counseling score

but again the stronger predictor was Factor E with 10.3% of the variance

and a simple correlation of .32. Factor D (Analytical Interest Vali-

dation) added 4.4% to the variance accounted for and raised the multiple

correlation to .38 with a total of 14.7% of explained score variance.

The second major hypothesis that supervisor ratings on the Coun-

selor Evaluation Rating Scale are predictable from HRPI response

preference scores, appears to be supported as far as Total score and

Counseling score are concerned. In this case only two HRPI subscores

are predicting the supervisor ratings, but attainment of the requi-

site .05 confidence level of prediction allows the conclusion that

Factors E and D did successfully predict the criterion, supporting the

second major hypothesis with the limitations which have been noted.

It is possible, therefore, to construct regression equations to

predict the seven scores of the two criterion instruments. Table 12

presents the regression equations for predicting BLRI scores and Table

13 shows the regression equations for the two CERS scores predictable

from the variables under study here. The two regression equation tables

culminate this examination of the predictive potential of the Helper

Response Preference Inventory vis-a-vis client and supervisory ratings.

















TABLE 12

REGRESSION EQUATION FOR THE BARRETT-LENNARD
RELATIONSHIP INVENTORY PREDICTORS


Dimensions


Equations


Total Score

Level of Regard


80.10

2.48

0.53


(GPA) -

(Factor

(Factor


44.75

19.08 (GPA) -


Empathy Level

Unconditionality
of Regard

Congruence


25.13

17.35


(GPA) -

(GPA) -


3.62 (Factor F) 154.76

J) + 13.69 (GPA) +

A) 0.98 (Factor I) -


1.12 (Factor F) 32.44


2.01 (Factor J) 56.58

31.91



















TABLE 13

REGRESSION EQUATION FOR THE COUNSELOR
EVALUATION RATING SCALE PREDICTORS


Equations


Total Score


2.32 (Factor E) + 140.10


Counseling Score


1,39 (Factor E) 0.84 (Factor D)


+ 65.94


Supervision Score


Sca I es


(No Predictors)

















CHAPTER V

DISCUSSION AND SU.IJM.1ARY


The review of research on counseling effectiveness in Chapter II

demonstrated the varied and often circuitous routes to building a con-

sistent model for selecting and training effective counselors. Never-

theless, the centrality and interrelatedness of the counselor's person-

ality and communication style emerge consistently across research as fun-

damental to the helping process. The literature review also suggested

that counselors may have characteristic manners of responding verbally

to clients. The propensity toward preferred responses is probably rooted

in both habit and value structure and may crucially shape the forma-

tion of the therapeutic relationship as well as the outcome of the

helping process.

This study was designed to investigate the relationship of response

choices on the Helper Response Preference Inventory by counselor

trainees to client perceptions of the relationship and supervisor

evaluations. Client perceptions were assessed through the Barrett-

Lennard Relationship Inventory (BLRI) and supervisor ratings were

collected by means of the Counselor Evaluation Rating Scale (CERS).

It was hypothesized, first of all, that a relationship would be

found between counselor trainee response preferences on the HRPI and

client evaluations on the BLRI. Second, it was hypothesized that a

relationship would be found between response preferences on the HRPI

79









and supervisor evaluations on the CERS. This study also investigated

the degree of relationship among the three complete sets of variables

produced by the two criterion instruments and the predictor instrument.

A group of 46 graduate-level counselor trainees participated in

all aspects of the study along with their clients and supervisors. The

trainees completed the HRPI and were in turn evaluated by their clients

on the BLRI and by their supervisors on the CERS. By means of a step-

wise regression technique the ten factor scores of the HRPI and four

selected demographic variables were regressed upon each of the criterion

instruments, employing both subscores and total scores. The findings

of this investigation were presented in Chapter IV.


Summary of the Results


Two directional hypotheses were tested during the investigation,

and each major hypothesis subsumed several hypotheses. The partial

questions break down each hypothesis into its components which are sub-

scale and total scale scores within the criterion instrument. The major

hypotheses, the attendant partial hypotheses, and the results of the

analytical process on each are:

HOI There is a relationship between counselor trainee response

preferences and client evaluations of their counseling effec-

tiveness. Fail to reject; partially.

I. A relationship exists between response preferences and Total

score on the BLRI. Fail to reject.

2. A relationship exists between response preferences and Level

of Regard score on the BLRI. Fail to reject.









3. A relationship exists between response preferences and Empathy

Level score on the BLRI. Fail to reject.

4. A relationship exists between response preferences and Uncon-

ditionality of Regard score on the BLRI. Fail to reject.

5. A relationship exists between response preferences and Con-

gruence score on the BLRI. Rejected.

HO2 There is a relationship between counselor trainee response

preferences and supervisor evaluations of their counseling

effectiveness. Fail to reject; partially.

I. A relationship exists between response preferences and Total

score on the CERS. Fail to reject.

2. A relationship exists between response preferences and

Counseling score on the CERS. Fail to reject.

3. A relationship exists between response preferences and Super-

vision score on the CERS. Rejected.


Practical Limitations


The nature of this research necessitates some consideration of the

actual limitations within which these findings must be interpreted.

In Chapter III it was pointed out that a deliberate sample limits

generalizability, that the external validity is compromised (somewhat)

by the exclusion of certain subjects, and that the primitive level of

development of the instruments involved makes dramatic assertions more

hazardous. In addition three more factors, in retrospect, may deserve

consideration as having influenced outcome.

An element of volunteerism was introduced when certain counselor

trainees chose not to participate in the research. The voluntary status










of involvement was accentuated by the failure of other counselor trainees

to follow through with participation to which they had earlier con-

sented. Volunteerism was further introduced when 15 clients failed to

return their rating scale for trainees whose data sets were otherwise

complete. Such non-participation, regardless of the cause, tends to

limit the generalizability of this investigation.

An artifact of the predictor instrument, the HRPI, is the restric-

tion of choice range. The forced-choice format precludes any measure-

ment of the creative capacity of the trainee to respond -in a helpful

way. Theoretically, there could be a myriad of possible responses which

could be assigned to a wide spectrum of potential helpfulness. The

ability to respond creatively may not be measured, and conceivably

negated, by the format of the predictor instrument.

A third possible practical limitation is simply a part of the real

world, and therefore a potential hazard for any research utilizing

rating scales. Examination of the data suggests that raters, both

clients and supervisors, may have been influenced by certain ingrained

response patterns themselves. For example, the CERS employs a six-

point scale with three positive and three negative responses. One might

speculate that the most compatible response would be the middle positive

one which implies exceeding the minimum requirements but leaving room

for improvement. If such a phenomenon operated for all raters, the

Counseling and Supervision mean scores on the CERS would have been 78

and the Total score 162. The observed scores were 81.33, 84.22, and

171.54 respectively. Apparently a rating set did not occur in some

cases. Nevertheless, a fine-grained sensitivity to the actual nuances

of difference among counselors may have been influenced by this very









human pattern. The presence of significant observed differences may be

more remarkable when considered in this light.


Interpretation and Discussion of Findings


Although the two major hypotheses could not be accepted in full,

there were certain discernible trends apparent in the findings. Clearly

the HRPI factors are predicting into two different spheres represented

by client perceptions of the relationship on one hand and supervisor

evaluations of the counselor's ability on the other. Six HRPI factors

achieved the .05 confidence level of significance as predictors. Four

predicted the BLRI scores and two predicted the CERS scores, but no

single factor predicted into both instruments.

The client perceptions registered on the BLRI appeared to place con-

siderable value upon self disclosure. Frank Feedback was the single most

potent predictor in this study for any individual criterion score, pre-

dicting the Level of Regard score. The same score also was predicted

in a lesser amount by responses from an external frame of reference which

in a sense supports the self disclosure theme. Along this same line the

BLRI Total score was predicted in part by premature judgments or quick

interpretations. The implication seems to be that clients perceive

honest, direct, and even potentially threatening feedback as concerned

involvement. In fact, in this study they experienced premature judg-

ments, as opposed to focused probing, as a higher level of empathy.

It is worth noting that the frank honesty valued by clients did

not appear to be unbridled bluntness. For instance, empathic under-

standing also was a significant predictor of Level of Regard. Despite

the potential threat of frank feedback from an external point of








reference, the communication style of the higher rated counselors

apparently was characterized by sensitivity and timing. The implica-

tion is that the counselor cared enough to experience with and share

openly concerning the harsh, sometimes painful, realities of life.

Nevertheless, the helper transmitted, despite the uncomfortable con-

tent, underlying caring for and commitment to the client.

It is interesting that Frank Feedback became a negative predictor

of the Unconditionality of Regard score. Perhaps this observation should

be interpreted in light of the growing disenchantment within counseling

theorist circles with the concept of unconditional regard. Some coun-

selors question the possibility or even the appropriateness of communica-

ting unconditional regard in the face of deteriorative or even destructive

client behavior. Certainly no definitive statement can be inferred from

this sparse evidence, but the field may continue to benefit from a

reexamination of the concept.

Unmistakably the most consistent predictor of BLRI scores was

graduate grade point average. This finding can be looked at in a number

of ways. In part the GPA had been influenced by practicum and intern-

ship grades so the counselor trainee's GPA reflected some "expert"

evaluation of his or her counseling ability as well as a component of

cumulative experience. In part higher GPA suggests a higher verbal

facility which may be an indirect measure of the core concern of this

study, communication style. Another plausible explanation of the rela-

tionship between high GPA and higher BLRI scores is a "prestige" factor.

Perhaps high GPA reflects an intelligence factor which impresses the

client. The counselor trainee with high GPA may be seen as one who is

knowledgeable, skilled, self-confident, and a worthwhile person with

whom to relate more closely.








A different pair of HRPI factors predicted two CERS scores. With

the supervisor criterion no demographic variable emerged as a predictor.

Also with the CERS, self disclosure did not appear to be meaningful

in the prediction process. The supervisor ratings were related to

clarification. Perhaps the supervisors rated positively the ability of

the counselor trainee to cut through superficial verbiage to highlight

implied meanings and lead the interchange into sharper focus. Clarifi-

cation was the only predictor of the Total score and by far the best

predictor of the Counseling score. The supervisors evidenced that they

value the trainee's skill in leading the client into deeper strata of

self understanding. This theme is supported by the only other predictor

of a CERS score. The Counseling score was also predicted by responses

that validate the client's experience. The implication is, "Yes, that's

the way you really are." The result is a very positive, affirming way

of developing the client's repertoire of personal insights.

Thus, in an overall perspective it can be said that the HRPI factor

scores predict both the BLRI and the CERS in divergent ways. For the

client data the response choices that favor frank self disclosure are

significantly predictive. For supervisor ratings it is those response

preferences that stress development of client self understanding which

predict at a meaningful level.


Application


The pertinent question that attends a study of this nature arose

in the introduction in Chapter I, persisted throughout Chapter II, and

in essence permeates this research. The question is whether the predic-

tor instrument under investigation can or should be used for counselor









selection, training, or skill analysis. These questions will be con-

sidered separately in light of the empirical findings.

The nature of a concurrent, cross-sectional investigation such as

this does not directly address the issue of counselor trainee selection.

To properly evaluate an instrument's usefulness as a screening tool in

the candidate selection process a longitudinal study of the instru-

ment's predictive validity is necessitated. Nevertheless in the real

world of practical affairs those persons charged with the responsi-

bility of candidate selection must base those decisions upon the best

available information. This instrument does provide some useful

information despite the modest variances accounted for, particularly

when used in conjunction with other indicators.

Perhaps the first question that must be resolved relates to which

definition of counseling effectiveness one wants to predict. Clearly

the HRPI predicts in different ways into both client and supervisor

definitions of effectiveness. Maybe the question is not an either-or

proposition since both criteria appear to be predictable from a single

test administration. Actually a candidate could be considered in the

light of his scores on the four factors which predict client ratings and

the two which predict supervisor ratings. One must do so, neverthe-

less, with full awareness that thus far the instrument has not been

investigated in a manner thoroughly consistent with such use.

Whether the HRPI should be used for training or skill analysis is

another matter. While the predictive power is much too low to make life

changing decisions on this basis, a counselor trainee may gain some

insight into his own communication style through the instrument, For

example, if the counselor trainee's profile reveals a preference away








from frank feedback or clarification, he may be alerted to ways he can

enhance his effectiveness in terms of his client's perception, his super-

visor's perception, or both. Likewise, he may discover that his candor

is not balanced by understanding or vice versa, sensitizing himself to

areas worthy of emphasis in his skill-building process. In short, the

Helper Response Preference Inventory shows promise, albeit tentative,

as a useful educational tool to help bridge the gap between self dis-

closure and empathic understanding and between the effectiveness models

of clients and supervisors.


Suggestions for Further Research


An examination of the findings of this study suggests that further

research in a number of areas may be beneficial. In general the ration-

ale that gave rise to this study and the literature review in Chapter II

speak clearly of the need for development of instruments designed to

measure counseling effectiveness. Specifically there is a need for

instruments to assess the communication style of relatively effective

counselors in a continuing effort to tease out the elements distinctively

characteristic of helpful communication.

Previous references have been made to the need for a longitudinal

study to evaluate the effectiveness over time of counselor trainees who

have taken the HRPI. The goal would be to arrive at a predictive formula

whereby a probability statement could be made regarding a counselor

trainee who attains a given score on the instrument. Of course, such

precision is a long way off and will doubtless prove to be elusive if

not unattainable. At the very least a thorough examination of how a

trainee's HRPI score profile changes over time, especially with

practical supervised training, could be highly instructive.







Research into the relationship of the HRPI to other instruments

used to evaluate counseling effectiveness as outlined in Chapter II may

be profitable. Specifically it may be productive to relate response

preferences to the Rokeach Dogmatism Scale, for example, or to any of

the personality measures such as the 16 Personality Factor Questionnaire

or the Edwards Personal Preference Schedule. It is conceivable that a

multiple criterion approach to counseling effectiveness might yield a

combination of personality variables plus communicational modes which

would produce a predictive formula robust enough to survive replication.

Finally, this present research study should be replicated. Repli-

cations should be carried out, first of all, under approximately equiva-

lent conditions to examine the reliability of these findings. Next it

would be useful to conduct similar investigations with differing samples

such as paraprofessionals and career helper trainees from various dis-

ciplines, representing a broad spectrum of theoretical orientations.

Perhaps it would be enlightening to carry out this research design using

the same predictor instrument but substituting different client and

supervisor ratings as criterion instruments. Clearly, this study and

parallel studies have hardly scratched the surface of what may prove

to be a fertile field of investigation.


Conclusions of the Investigation


With due regard for the possible limitations of the study cited

in Chapter III and expanded in this present chapter, there are certain

tentative conclusions which can be drawn. First, the Helper Response

Preference Inventory does appear to predict client perceptions of the

counseling relationship on the BLRI in the case of four factors which








explain relatively modest portions of the score variance. Secondly,

the HRPI appears to predict supervisor ratings of the counselor trainee

on the CERS with two factors at a slightly more modest level. Thirdly,

the BLRI and the CERS appear to be measuring different aspects of the

counseling arrangement, and consequently the HRPI predicts into each

criterion in different ways with different response factors. Subse-

quent use of the HRPI should be made in the light of these findings,

and future investigations in turn may amplify or modify the signifi-

cance of these results.




































APPENDICES















APPENDIX A


COUNSELOR TRAINEE DEMOGRAPHIC DATA SHEET


Your name:_____

Phone number:____

Sex:

Age:

Undergraduate major:


Cumulative graduate grade point


average at the end of last quarter:


Number of graduate quarters you will have completed at the end of this
quarter:

Number of quarters of supervised counseling practice (practicum or
internship) you will have completed at the end of this quarter:


Individual supervisor:

Present practicum or internship agency:




Full Text
73
TABLE 10
BARRETT-LENNARD RELATIONSHIP INVENTORY PREDICTORS
Criterion Predictors
Multiple R
R2
df
F
Total Score
Graduate Grade Point Average
0.35
0.13
1,44
6.34*
Factor F (Focused Probing -
Premature Judgments)
0.42
0. 18
2,43
4.61*
Level of Reqard
Factor J (Frank Feedback -
1ncongruence)
0.41
0.17
1 ,44
8.70*
Graduate Grade Point Average
0.50
0.25
2,43
7.06*
Factor A (Empathic Under
standing Probing)
0.53
0.29
3,42
5.58*
Factor 1 (Internal External
Frame of Reference)
0.57
0.32
4,41
4.83*
Empathy Level
Graduate Grade Point Average
0.28
0.08
1 ,44
3.72*
Factor F (Focused Probing -
Premature Judgments)
0.36
0.13
2,43
3.28*
Unconditionality of Reqard
Graduate Grade Point Average
0.33
0. 1 1
1 ,44
5.44*
Factor J (Frank Feedback -
1ncongruence)
0.41
0. 17
2,43
4.42*
Conqruence
Graduate Grade Point Average
0.30
0.09
1,44
4.27*
*p<.05


102
Helper A:
"I'm sure I'll be able to help you."
Helper B:
"You're pretty concerned about status, aren't you?"
5. Helpee:
"1 feel so good, lately, now that I'm beginning to see
how I've been bringing some of these things on myself."
Helper A:
"You seem to be enjoying 1 ife a little more, now that
you feel that you understand how some of your diffi
culties have been coming about."
Helper B:
"Life really begins to mean something when you face
your responsibilities, doesn't it?"
6. Helpee:
"1 don't see how you can manage to sif and listen to me
week after week without being bored out of your mind."
Helper A:
"But you really don't bore me at all."
Helper B:
"You seem to feel that there's nothing about you that
anyone could find interesting."
7. Helpee:
"My scores on all of the tests I've taken so far point
toward success in sales work, but every time 1 just
about get up enough courage to switch majors, 1 get cold
feet and back out."
Helper A:
"You seem to be drawn to selling becuase of the rewards
you might realize there, but some of the interpersonal
skills required are causing you to re-evaluate your
decisi on."
Helper B:
"Logically, you have some evidence that this is some
thing you could do, but somehow you feel that there's
something about this that isn't quite right for you."
8. Helpee:
"My friends get all upset just because I'm a pretty
heavy drinker. 1 don't see how that should concern
them, just so long as 1 don't get in their way."
Helper A:
"It's really up to you, isn't it? If it bothers them,
that's really their problem, so long as you can handle
it."
He 1 per B:
"But you can't live your life alone. You ought to think
of how you're affecting other people through your
behavior."
9. Helpee:
"1 just feel so alone right now, just like I'm back
inside a shell of some kind with no way to reach out."
Helper A:
"But you really aren't alone, and you can reach out.
People will respond."


50
upon which the instrument was based established a defensible justifica
tion for utilizing the Relationship Inventory as a criterion in this
study.
The basic theory underlying the Relationship Inventory (Rogers,
1957; Barrett-Lennard, 1962) has been substantially supported by the
vast range of studies using the Truax and Carkhuff rating scales to
measure levels of therapeutic conditions within the context of relation
ship (Truax & Carkhuff, 1964, 1967; Carkhuff & Truax, 1965a, 1965b;
Berenson, Carkhuff, & Myrus, 1965).
More pointed research has involved the instrument per se. Success
ful use of the Relationship Inventory to appraise counselor-client rela
tionships has been reported by Van der Veer (1961), Barrett-Lennard
(1962), Cahoon (1962), and Gross and Deridder (1966). Other intimate
relationships measured with significant results include marital partners
(Thornton, I960), student-parent (Hollenbeck, 1965), supervisor-teacher
(Blumberg, 1968), and student-teacher (Mason and Blumberg, 1969).
Barrington (1961) related scores on the Barrett-Lennard Relation
ship Inventory to such specific counselor behaviors as a decrease in
the length of counselor responses from the first to the second inter
view, percentage of emotional words used by the counselor, and simi
larity of the counselor's rate of speech to the client's rate. The most
predictive sub-scale was the empathic understanding dimension, but
perceived counselor congruence was highly related to more integrated
verbalizing by the counselor in the second hour of counseling.
Using a Q-sort technique, Emmerling (1961) studied the relation
ship of "openness" in relatively effective and ineffective teachers to
how their students perceived them in relationship. It was hypothesized


40
requested to administer the Barrett-Lennard Relationship Inventory to
one of their clients. The trainees were then informed about the adminis
tration procedure. Specific verbal instructions to the counselor
trainees were as follows:
From among the clients you have seen in a counseling relationship
for a minimum of two thirty-minute sessions, select your most
typical client. Ask the selected client to please complete the
Barrett-Lennard Relationship Inventory based upon your counseling
contacts together. Make sure the client reads the cover letter
of explanation. Also, be certain that the client thoroughly under
stands the rating procedure. Emphasize that he or she is the only
client you are asking so cooperation is extremely important.
Offer to stay with the client while the first page of the Rela
tionship Inventory is completed in private, to answer any questions
that might arise. When you are convinced the client understands
the task and will follow through, stress the fact that the instru
ment will be ready to mail as soon as completed and leave the client
to finish the ratings.
The client was then expected to comply with the verbal instructions
to evaluate and mail the instrument. Similar instructions were also
reinforced in the explanatory cover letter (see Appendix). The instru
ment was stamped and pre-addressed so the client could personally mail
the Inventory to the researcher. The brief cover letter, in an effort
to maximize the probability of genuinely candid responses, explained the
procedure in part. It was emphasized that even though the ratings were
important, they would not be used to personally evaluate the individual
counselor trainee. Also stressed was the fact that since all information


CHAPTER I I
REVIEW OF RELATED RESEARCH
The review of literature focuses on the following areas: I) the
varied approaches to conceptualization and measurement of counselor
effectiveness, and 2) the application of supervisors' and clients'
ratings to counselor effectiveness research.
Approaches to Counselor Effectiveness
The pursuit of a consistently viable way to assess the effective
counselor has often led to confusing and contradictory results. Quite
early, the sentiment had crystallized, and was later traced historically
by Polmantier (1966), that the profession needed to recruit counselor
candidates with personal qualities which would enhance their success as
counselors. Subsequently, the process of identification of those
essential personal qualities has met with mixed results. In this review,
five primary approaches to the search for a definition of counselor
effectiveness are explored: The speculative-philosophical approach, the
psychometric approach, the communication analysis approach, the percep
tual approach, and the core dimensional approach derived from process
research. After the first two approaches are presented, the question of
the criterion problem in effectiveness research is discussed to intro
duce the consideration of the final three approaches to counselor
effectiveness research.
7


106
Helper A:
"Perhaps there was more conflict in your family than you
realized at the time."
Helper B:
"Its good that youve recognized the trouble before it
got too serious."
28. Helpee:
"Every day 1 go to class, 1 get this panicky feeling,
just like I'm losing my mind or something like that."
Helper A:
"Tell me, if you can, where you get this feeling most
often?"
Helper B:
"This must be pretty frightening for you."
29. Helpee:
"After the depression, 1 never got back on my feet
financially, and it seems that 1 just never could make
ends meet after that."
Helper A:
"So things have been tough for quite a long time."
He 1 per B:
"Well, let's hope things will be better shortly."
30. Helpee:
"When 1 left the state university last year, 1 was real
confident that 1 could make up my mind about what 1
wanted to do. But the longer I'm out, the more confused
things seem."
Helper A:
"At the time you left, what specific plans had you
formu1ated?"
Helper B:
"It often takes some time to really decide what you do
want to do, especially right after you finish college."
31. Helpee:
"1 don't believe much in religion any more. I've
gradually come to realize that most people simply use
religion as a crutch to compensate for their own weak
nesses."
Helper A:
"But I'm sure you know that some strong persons really
do have deep religious convictions."
Helper B:
"Do you mean that religion hasn't really been of help
to you, that you've had to work things out for yourself?"
32. Helpee:
"1 did well on the college entrance exam and the reading
test, but my math aptitude scores were way below average.
Do you think that wi11 cause problems for me?"
Helper A:
"1 don't think so unless you plan a program requiring a
lot of math. This suggests that you've either had little
work in math or that you really didn't find it very
cha 1lenging."


and Wargo (1966) were found to have higher scores on "change" and
"autonomy" and lower scores on "order." Oh I sen (1967), also using the
Edwards Personal Preference Schedule, measured two separate sample groups
of counselors. In the first group the most effective counselors scored
higher than the least effective on "succorance" and lower on "order,"
"deference," and "consistency." In contrast, Ohlsens most effective
counselors in the second group scored higher than the least effective
on "intraception" and lower on "dominance" and "aggression." Demos and
Zuwaylif (1966) found the EPPS to be the only one of three psychometric
measures to discriminate accurately among 30 secondary school counselors
in an NDEA institute based upon "multiple objective and subjective
criteria." Their most successful counselors scored significantly higher
on "nurturance" and "affiliation" while the least successful group
scored significantly higher on "autonomy," "abasement," and "aggression."
It is apparent that despite the similarities overall, the general
pattern of results is much to inconsistent to permit the broad scale
use of the EPPS as a reliable predictor of effective counseling.
Another standardized personality instrument which has been applied
in attempts to measure counselor effectiveness is the 16 Personality
Factor Profile (16 PF). In 1968 a regression equation was derived by
McClain for predicting counselor effectiveness from 16 PF scores. The
limited use of this prediction equation may be related to the generally
significant but inconsistent results found in early research attempts
to use the 16 PF to differentiate levels of effectiveness (Donnan,
Harlan and Thompson, 1969; McClain, 1968; Myrick, Kelly, & Wittmer,
1971; and Wittmer & Lister, 1971).
The Rokeach Dogmatism Scale has been used with considerable promise


8
The Speculative-Philosophical Model
The most primitive approach to establishing the components of
effective counseling began with the emergence of serious attempts to
conceptualize the necessary ingredients. In the widely quoted Fiedler
(1950) study it was concluded that experienced counselors, regardless
of frame of reference, agreed on the fundamental characteristies of an
ideal counselor-client relationship. The counselors concurred that the
ideal therapeutic relationship involves I) an empathic relationship,
2) rapport between client and counselor, 3) the counselor staying with
the client's problems, 4) the freedom to self-disclose, 5) mutual trust
and confidence, 6) excellent rapport, 7) active client participation,
8) freedom to make choices by client, 9) acceptance of client feelings
as normal and understandable, 10) a tolerant atmosphere, and II) a
feeling of being understood on the client's part.
Concurrent with the analysis of the ideal relationship was a
growing interest in those personal qualities of the counselor which
engender and enhance this therapeutic relationship. In their review of
this early literature, Barry and Wolf (1958) concluded that the
philosophical approach in this era consisted primarily of cataloguing
assorted virtues to be sought in prospective counselors. Graves (1944)
specified integrity, vitality, judgment, industriousness, high personal
standards, adaptability, experience and training. Cox (1945) added such
qualities as fairness, sincerity, good character and wholesome philosophy.
In addition to the above, Bowler and Dawson (1948) emphasized maturity,
respect, resourcefulness, reliability, a sense of humor, and an ability
to listen and keep confidences. Then in 1949 the National Vocational
Guidance Association proclaimed that counselors, ideally, were interested


THE RELATIONSHIP BETWEEN RESPONSE PREFERENCES
AND COUNSELOR EFFECTIVENESS RATINGS
OF COUNSELOR TRAINEES
By
SIDNEY LEON BRADLEY
A DISSERTATION PRESENTED TO THE GRADUATE COUNC
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR
DEGREE OF DOCTOR OF PH ILOSOPHY
L OF
THE
UNIVERSITY OF FLORIDA
1976


124
Rosen, A. Client preferences: An overview of the Iiterature. Personnel
and Guidance Journal, 1967, 45, 785-789.
Rotter, J. C. The perceptuaI character istics of elementary, secondary,
and community college counselors. Doctoral dissertation, Wayne
State University, 1971.
Rushong, H. D. Present status and trends in the evaluation of counsel
ing. Educational and Psychological Measurement, 1953, 15, 418-430.
Russo, J. R., Kelz, J. W., & Hudson, G. R. Are good counselors open-
minded? Counselor Education and Supervision, 1964, _3, 74-77.
Satir, V. Conjoint Family Therapy. Palo Alto: Science and Behavior
Books, 1967.
Schmidt, G. D. Faculty perceptions of clergy in counselor education
programs. Counselor Education and Supervision, 1974, 13, 294-298.
Schroeder, P., & Dowse, E. Selection, function and assessment of
residence hall counselors. Personnel and Guidance Journal, 1968,
47_, 151-156.
Severinsen, K. N. Client expectations and perception of the counselor's
role and their relationship to client satisfaction. Journal of
Counseling Psychology, 1966, 13, 109-112.
Shapiro, J. G. Relationships between expert and neophyte ratings of
therapeutic conditions. Journal of Consulting and Clinical Psychol-
ogy, 1968, 32, 87-89.
Shoben, E. J. Research in counseling: A symposium; some problems in
establishing criteria of effectiveness. Personnel and Guidance
Journal 1953, 2U_, 284-294.
Snelbecker, G. E. Influence of therapeutic technigues on college
students' perceptions of therapists. Journal of Consulting Psychol
ogy, 1967, 31, 614-618.
Sonne, T. R., & Goldman, L. Preferences of authoratarian and eguali-
tarian personalities for client-centered and selectic counseling.
Journal of Counseling Psychology, 1957, £, 129-135.
Stablein, J. E. A critigue of counseling evaluation. Personnel and
Guidance Journal 1962, £l_, 66-68.
Steffi re, B., King, P., & Leafgren, F, Characteristics of counselors
judged effective by their peers. Journal of Counseling Psychology,
1962, 9, 335-340.
Stern, G. G., Stain, M. L., & Bloom, B. S. Methods in Personality
Assessment. Glencoe: Free Press, 1956.


104
Helper A:
"How do you know that this cheating is going on?"
Helper B:
"You're probably afraid to put forth real effort for
fear of failure."
16. Helpee:
"Im really surprised at these test results! 1 never
expected to have such a high scientific interest score."
Helper A:
"Why do these results surprise you so much?"
Helper B:
"Yes, your score is quite high, very consistent with
your good grades in physics and chemistry."
17. Helpee:
"I've never felt comfortable when I'm with people. Now
that I'm having to do more public speaking, it's really
getting to me."
Helper A:
"It's getting pretty frustrating, isn't it?"
Helper B:
"It's probably worse now that you're in a position where
people may be evaluating you."
18. Helpee:
"I've thought about the pros and cons, and 1 still don't
know which way 1 should go. What do you think 1 should
do?"
Helper A:
"You should be depending on your own judgment, not mine."
Helper B:
"You seem uneasy about facing the consequences of either
course of action."
19. Helpee:
"My husband is a fine person and good to the children,
so 1 guess it isn't fair for me to criticize him."
Helper A:
"You dislike some things about him, but feel guilty
about it."
Helper B:
"What do you think would happen if you were to express
some of your feelings to him?
20. Helpee:
"I've thought a lot about how 1 affect other people. 1
know that I've had a lot of trouble, but it's hard to
see just how 1 bring it on myself, if that's what really
happens."
Helper A:
"But you should examine your own behavior closely."
Helper B:
"You feel that you really aren't at fault much of the
time?"
21. Helpee:
"1 won't sit down! All we ever do is sit and listen
while you talk to us!"


108
38. Helpee:
"My brother was a lot older than 1, and we never had a
lot in common, but 1 always looked up to him, a lot more
than 1 did to my father."
Helper A:
"How did you get along with your father?"
Helper B:
"Maybe your brother really acted as a father toward
you."
39. Helpee:
"1 want to take my comprehensive exams this term, but
1 may decide to wait. Some of the people writing this
time are the real top students, and 1 might have a
better chance to pass next time."
Helper A:
"But you've prepared well for the exams and have a good
chance of passing, probably better than you give your-
self credit for."
Helper B:
"1 think you may be avoiding really finding out what
you are able to do on the exams."
40. Helpee:
"We've talked several times now, but things don't seem
to be changing much. Do you think we're getting any
where?"
Helper A:
"But this is the way we have to continue in order for
you to make the kind of progress you want."
Helper B:
"Are you beginning to wonder if this is a waste of time?"
41. Helpee:
"When is mommie coming to get me?"
Helper A:
"I'm sure she'll be here very soon."
Helper B:
"It'll be good when she gets here, won't it?"
42. Helpee:
"1 haven't taken anything from a store now for almost
two months. I'm hoping this shoplifing binge is over
for good."
Helper A:
"Have you had the urge to pick up things the last few
times you've been shopping?"
Helper B:
"That's a real improvement. I'm glad for you."
43. Helpee:
"Mr. Smith, you don't know how surprised 1 was when 1
found out that I'd passed your course. 1 know my grades
were awful, and 1 was almost positive that Id goofed
up the final exam."
Helper A:
"Were you surprised because you thought you were failing
up to the point of the exam?"


APPENDICES


Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy
THE RELATIONSHIP BETWEEN RESPONSE PREFERENCES
AND COUNSELOR EFFECTIVENESS RATINGS
OF COUNSELOR TRAINEES
By
Sidney Leon Bradley
December, 1976
Chairman: Joseph Wittmer
Major Department: Counselor Education
This study investigated the relationship between response preferences
of counselor trainees and effectiveness ratings by their supervisors and
clients. A total of forty-six counselor trainees, enrolled in a graduate-
level practicum or internship, participated as subjects in this research.
Each counselor trainee completed the predictor instrument, the Helper
Response Preference Inventory (HRPI), which yields ten factor-analyzed
sub-scales. Client perceptions of the relationship were obtained by
having the trainee rated by a chosen client on the Barrett-Lennard Rela
tionship Inventory (BLRI). The supervisor criterion instrument was the
Counselor Evaluation Rating Scale (CERS) on which each trainee was
evaluated by an individual supervisor.
It was hypothesized that a relationship would be found between
response preferences on the HRPI and client evaluations on the BLRI.
It was also hypothesized that a relationship would be found between HRPI
responses and supervisor evaluations on the CERS. The forward stepwise
vu


81
3. A relationship exists between response preferences and Empathy
Level score on the BLRI. Fail to reject.
4. A relationship exists between response preferences and Uncon
ditionality of Regard score on the BLRI. Fail to reject.
5. A relationship exists between response preferences and Con
gruence score on the BLRI. Rejected.
HC>2 There is a relationship between counselor trainee response
preferences and supervisor evaluations of their counseling
effectiveness. Fail to reject; partially.
1. A relationship exists between response preferences and Total
score on the CERS. Fail to reject.
2. A relationship exists between response preferences and
Counseling score on the CERS. Fail to reject.
3. A relationship exists between response preferences and Super
vision score on the CERS. Rejected.
Practical Limitations
The nature of this research necessitates some consideration of the
actual limitations within which these findings must be interpreted.
In Chapter III it was pointed out that a deliberate sample limits
genera IizabiIity, that the external validity is compromised (somewhat)
by the exclusion of certain subjects, and that the primitive level of
development of the instruments involved makes dramatic assertions more
hazardous. In addition three more factors, in retrospect, may deserve
consideration as having influenced outcome.
An element of volunteerism was introduced when certain counselor
trainees chose not to participate in the research. The voluntary status


82
of involvement was accentuated by the failure of other counselor trainees
to follow through with participation to which they had earlier con
sented. Volunteerism was further introduced when 15 clients failed to
return their rating scale for trainees whose data sets were otherwise
complete. Such non-participation, regardless of the cause, tends to
limit the genera I izabi Iity of this investigation.
An artifact of the predictor instrument, the HRPI, is the restric
tion of choice range. The forced-choice format precludes any measure
ment of the creative capacity of the trainee to respond in a helpful
way. Theoretically, there could be a myriad of possible responses which
could be assigned to a wide spectrum of potential helpfulness. The
ability to respond creatively may not be measured, and conceivably
negated, by the format of the predictor instrument.
A third possible practical limitation is simply a part of the real
world, and therefore a potential hazard for any research utilizing
rating scales. Examination of the data suggests that raters, both
clients and supervisors, may have been influenced by certain ingrained
response patterns themselves. For example, the CERS employs a six-
point scale with three positive and three negative responses. One might
speculate that the most compatible response would be the middle positive
one which implies exceeding the minimum requirements but leaving room
for improvement. If such a phenomenon operated for all raters, the
Counseling and Supervision mean scores on the CERS would have been 78
and the Total score 162. The observed scores were 81.33, 84.22, and
171.54 respectively. Apparently a rating set did not occur in some
cases. Nevertheless, a fine-grained sensitivity to the actual nuances
of difference among counselors may have been influenced by this very


CHAPTER I
INTRODUCTION
Purpose of the Study
The measurement of counseling effectiveness and prediction of
counseling'success are both central concerns to the counseling pro
fession in the era of accountability. Efficiency and professional
responsibility dictate the need for valid ways to evaluate competence
and to differentiate levels of effectiveness. This study has explored
the relationship between the effectiveness ratings of counselor trainees
and their characteristic ways of responding to clients as a possible
method of evaluating effectiveness. More generally, this study has
investigated the theory that a counselor's communication within the
helping relationship is related to his level of effectiveness as a
helping person. This study also attempted to specify measurable compo
nents of communication which distinguished more effective counselor
trainees in a graduate level training program from less effective
counselor trainees.
Rationale for the Study
Research that leads to identification of effective counseling is
clearly relevant to the process of counselor selection (ACES, 1967),
evaluation (Stablein, 1962), and prediction of future counseling behav
ior (Walton & Sweeney, 1969). Nevertheless, investigations which have
attempted to specify and predict effective counseling have frequently
I


10
important. Consequently, the pursuit ot measurable personality
variables of effective counselors began in earnest using more objective
measures.
The Psychometric Model
The use of psychometric test data to link personal variables with
counseling effectiveness has been a popular, broad-based movement.
Actually a number of researchers have been able to effectively discrimi
nate quality levels in counseling using personality measures. However,
in general, the differentiations between more and less effective
counselors have not succeeded in identifying a dependable instrument
for screening. The principal reasons why such an instrument has not
been found are that positive findings are seldom replicated and results
add up to be in consistent and inconclusive. In general, the psycho
metric approach has encompassed personality, interest, and attitude
measures of counselor effectiveness. To simplify the pro Iiferation of
research in this area, this review will examine several popular measures,
one at a time.
The Edwards Personal Preference Schedule (EPPS) has been widely
used as a personality instrument to discriminate effective from less
effective counselors. However, the results overall have been per
plexing. Steffi re, King, and Leafgren (1962) used the Edwards Personal
Preference Schedule to examine participants in a one semester institute.
They found "most chosen" counselors, on the basis of potential, scored
higher on "deference" and "order" and lower on "aggression" and "abase
ment" than the "least chosen" counselors. On the other hand, the
counselors judged to be more effective in the study by Truax, Silber,


APPENDIX F
HELPER RESPONSE PREFERENCE INVENTORY
Pi reef ions: This instrument is designed to measure your preference for
various types of responses which one person (the helper) might reasonably
make to another (the helpee) in an effort to assist him to solve his
problems. The Inventory consists of typical helpee statements each
followed by a pair of possible helper responses. For each helpee state
ment, select the helper response which you believe would be most appro
priate.
Be sure to select only one response for each helper
statement.
Do not leave any item blank.
You should assume that each of the helpee statements
was made by a different person.
Mark your preference on the answer sheet provided.
Do not mark or write on the inventory.
I.Helpee: "I think I really should talk with someone about my
problems, but I'm not so sure that I can."
Helper A; "What kind of problems do you seem to be having?"
Helper B: "You really shouldn't be upset about discussing your
problems. Everyone has difficulties."
2.Helpee: "No matter how hard I study, I seem to keep making silly,
stupid mistakes in math!"
Helper A: "What kind of grades have you been getting in your math
course?"
Helper B: "This must be pretty frustrating for you after working
at it so."
3.Helpee: "My wife and I were getting along pretty well until her
first pregnancy, but things have been pretty much going
downhill since then."
Helper A: "It sounds like you may have resented her pregnancy."
Helper B: "That isn't an unusual reaction to the first pregnancy."
4.Helpee: "I really had expected to see a psychiatrist when I
made this appointment. You're a social worker, aren't
you?"
101


58
Preference Inventory and the four demographic factors were predictor
variables regressed upon each of the dependent criterion variables. The
stepwise solution program (Kerlinger, 1973) allows at each step the
variable that makes'the greatest increment to the multiple coefficient
2
of determination, R to enter the computational process and calculates
a regression equation for that step. Also at each step of the analysis,
several calculations are performed and printed. These include the mul
tiple correlation coefficient, the regression equation, partial correla
tions, and the regression and residual sums of squares. One advantage
of this stepwise program is that the investigator can specify the level
of F ratio for entering a variable into the equation and the level of
F for removing a variable from the equation. For this study an alpha
level of .05 was considered the level of significance necessary to fail
to reject the hypotheses.
Utilizing the stepwise regression procedure made it possible to
ascertain which combination of independent variables, HRPI scores,
demographic data or combinations thereof, was the most efficacious pre
dictor of the dependent criterion variables, the Relationship Inventory
or the Counselor Evaluation Rating Scale. The regression equation
computed specified the precise beta weights for the independent variables
which, taken in combination, yielded the greatest reduction in the error
sums of squares when predicting either of the criterion variables.
2
This method of analysis also permitted an examination of the R s
and F ratios produced by each variables increase in the overall pre
diction process. By observing the increment to prediction for each
variable, the relative efficacy of different variables in the regression
equation became apparent. In effect, each hypothesis was tested for the


30
criteria. Both supervisor and client ratings are explored in some depth
in the review which follows.
Supervisor Ratings
There appears to be a general consensus that supervisor ratings are
among the most valid measures of effective counseling (Patterson, 1964).
The assumption is made that highly experienced counselors who comprise
"expert" panels are the most qualified of all judges to identify differ
ential levels of effective counseling. The logic of Cattell, penned in
1903, is still compelling today: "There is, however, no other criterion
for a mans work than the estimation in which it is held by those most
competent to judge" (p. 314). Clearly the determination of "those most
competent to judge" remains a debatable point, but Cattells rationale
continues to be invoked to justify the use of supervisor ratings.
Supervisor ratings have been frequently employed as a criterion
measure of counselor effectiveness. Such ratings were used in studies
by Allen (1967), Bishop (1971), Brown and Cannaday (1969), Demos and
Zuwaylif (1966), Embree (1954), Friesen and Dunning (1973), Gade (1967),
Keltz (1966), McDougalI and Reitan (1961), Watley (1967), Wicas and
Mahan (1966), and a host of others. For the most part interrater relia
bilities in these studies have been within the acceptable range. We
note, for example, that in the study by Gade (1967) the interstaff rank
ing achieved a reliability of .80 when computed by the coefficient of
concordance statistical formula. In their 1973 study Friesen and
Dunning found that supervisors had an interrater reliability of .94
on the Counselor Effectiveness Scale. This coefficient was higher than
the .92 achieved by lay people or the .91 reliability among practicum


56
eight personality factors with score differences between the groups
which were statistically significant.
In a challenge>of the previous research, Jones (1974) questioned
the appropriateness of using an instrument designed to help supervisors
evaluate practicum students behavior as a criterion of counseling
effectiveness. He argued, "To be valid as a criterion measure of coun
selor effectiveness, an instrument such as the CERS must be related to
counseling outcome or some other aspect of job performance" (p. 113).
Following his line of reasoning, Jones attempted to investigate the
relationships between the CERS and empathy, genuineness, and respect
using the Carkhuff Scales since these variables "have been established
as being significantly related to counseling outcome" (p. 113). He
found a significant positive relationship only between the level of
respect offered in the counseling situation and supervisor ratings on
the S (supervision) scale of the CERS. He concluded:
. . there appears to be no relationship between levels
of empathy, respect, and genuineness and the other CERS
scales, although the positive correlations between respect
and the CERS scales of C (counseling) and T (total) may
be noteworthy. (p. 115)
Forster and Hamburg (1976) in a recent study of 30 full-time coun
selor trainees failed to find a significant relationship between CERS
ratings and 16 PF scores.
A very recent study by Loesch and Rucker (1976), which has been
submitted for publication, has provided some support for the original
CERS Scales. In a second order factor analysis of more than four
hundred sets of ratings they obtained an oblique solution which closely
approximates the original classification of items into supervisory and
counsel ing sea Ies.


120
Klein, D. F., & Cleary, T. A. Platonic true scores and error in
psychiatric rating scales. Psychological Bulletin, 1967, 68, 77-80.
Knupfer, G., Jackson, D. D., & Krieger, G. Personality differences
between more and less competent psychotherapists as a function of
criteria of competence. Journal of Nerve and Mental Disorders,
1959, 129, 375-384.
Krumboltz, J. D. (Ed.) Revolution in Counselings: Imp Iications of
Behavioral Science. Boston: Houghton-M¡ffIin, 1966.
Kurtz, R. R., & Grummon, D. L. Different approaches to the measurement
of therapist empathy and their relationship to therapy outcomes.
Journal of Consulting and Clinical Psychology, 1972, 39, 106-115.
Lanning, W. L. A study of the relation between group and individual
counseling supervision and three relationship measures. Journal of
Counseling Psychology, 1971, J_8, 401-406.
Lanning, W. L., & Lemons, S. L. Another look at the factor structure of
the Barrett-Lennard Relationship Inventory. Measurement and Evalua
tion in Guidance, 1974, 6_, 228-231.
Lawlis, G. F., & Lu, E. Judgment of counseling process: Reliability,
agreement, and error. Psychological Bui let in, 1972, 78, 17-20.
Lesser, W. M. The relationship between counseling progress and empathic
understanding. Journal of Counseling Psychology, 1961, (3, 330-336.
Lin, T. T. Counseling relationship as a function of counselors self
confidence. Journal of Counseling Psychology, 1973, 20, 293-297.
Linden, J. D., Stone, S. C., & Shertzer, B. Development and evaluation
of an inventory for rating counseling. Personnel and Guidance
Journal. 1965, 44,267-276.
Lister, J. L. Reliability of the Helper Response Preference Inventory.
Unpublished mimeograph report, University of Florida, 1972.
Lister, J. L. School counseling: For better or for worse? Caadian
Counselor. 1970, 4_, 33-39.
Loesch, L. C. NonverbaIized feelings and perceptions of the counseling
relationship. Counselor Education and Supervision, 1975. 15. 105
113.
Loesch, L. C., & Rucker, B. B. Another view on the validity of the
Counselor Evaluation Rating Scale (submitted for publication).
Lou, M., & McNair, D. M. Methods relating to evaluation of therapeutic
outcome. In Gottschalk, L. A., & Auerback, A. H. (Eds.), Methods
of Research in Psychotherapy. New York: AppIeton-Century-Crofts,
1966.


76
Superficiality) which accounted for 7.4$ of the variance. With only
one predictor the simple correlation and the multiple correlation were
an identical .27.
There were two factors predicting into the CERS Counseling score
but again the stronger predictor was Factor E with 10.3$ of the variance
and a simple correlation of .32. Factor D (Analytical Interest Vali
dation) added 4.4$ to the variance accounted for and raised the multiple
correlation to .38 with a total of 14.7$ of explained score variance.
The second major hypothesis that supervisor ratings on the Coun
selor Evaluation Rating Scale are predictable from HRPI response
preference scores, appears to be supported as far as Total score and
Counseling score are concerned. In this case only two HRPI subscores
are predicting the supervisor ratings, but attainment of the requi
site .05 confidence level of prediction allows the conclusion that
Factors E and D did successfully predict the criterion, supporting the
second major hypothesis with the limitations which have been noted.
It is possible, therefore, to construct regression equations to
predict the seven scores of the two criterion instruments. Table 12
presents the regression equations for predicting BLRI scores and Table
13 shows the regression equations for the two CERS scores predictable
from the variables under study here. The two regression equation tables
culminate this examination of the predictive potential of the Helper
Response Preference Inventory vis-a-vis client and supervisory ratings.


61
Relationship Inventory. Of those 67 counselor trainees, two chose not
to participate for personal reasons, one was unavailable during the data
collection period, and three failed to return even the demographic data
sheet and HRPI answer sheet after agreeing to do so, despite follow
up attempts. At least partial data were compiled on 61 counselor
trainees. Supervisors' ratings were obtained on all 61 subjects in
addition to the demographic data and the HRPI answer form. However,
the client rating on the Relationship Inventory was returned on only
46 counselor trainees. Therefore, the net sample for this study was
46 counselor trainees.
Descriptively, the counselor trainee sample was composed of 27
males and 19 females. The age range was 22 to 51 years, and the mean
age was 27.6. The graduate grade point average ranged from 3.20 to 4.00
with a mean of 3.77. The counselor trainees in the study had completed
from 3 to 17 graduate quarters and I to 8 quarters of supervised prac
tice with respective means of 7.6 and 4.1. A breakdown of demographic
data on the counselor trainees is presented in Table 4.
The counselor trainees came from a wide variety of undergraduate
major backgrounds. Although the majority of trainees (27) had chosen
psychology as an undergraduate major, eight additional fields were
represented. There were five trainees each from education and language
undergraduate programs. Likewise sociology, religion, and the physical
sciences were represented by two counselor trainees each. There was
one counselor trainee each from political science, art history, and
business. The overall impression is that the sample represents a
diversity of backgrounds, interests, and personalities.


regression technique was used to analyze the data. Four demographic
variables: age, graduate grade point average (GPA), number of graduate
quarters completed, and number of quarters of supervised counseling
practica completed, were also entered into the regression equations.
The best predictor of the BLRI Total score was grade point average
followed by responses that were premature judgments. The BLRI Level of
Regard subscore was predicted best by frank feedback responses. The
second best predictor was GPA followed by empathic responses and then
responses from an external frame of reference. The BLRI Empathy Level
subscore was predicted best by GPA, but the scale was also predicted by
premature judgments. For the BLRI Unconditionality of Regard subscore
GPA was the strongest predictor, but frank feedback was a significant
negative predictor. Only GPA predicted the BLRI Congruence subscore.
Other than GPA the demographic variables were not significant predictors.
The only statistically significant predictor of the CERS Total score
was the set of responses which used clarification primarily. The
clarification response factor also was the best predictor of the
Counseling subscore of the CERS although the scale was also predicted
by responses that validate or affirm the counselee's identity. There
was no predictor of the CERS Supervision subscore. No demographic
variable predicted any of the CERS scores.
It was noted that no HRPI factor predicted successfully both BLRI
and CERS scores. Also the canonical correlation between any two of the
three instruments in the study was not found to be significant. Inter-
correlations between the instruments were also generally low. Never
theless, neither hypothesis, that a relationship exists between the HRPI
and each of the two criterion instruments, could be rejected except in
part.
viii


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 EO1HXQ71A_VVGLP0 INGEST_TIME 2014-10-06T22:52:25Z PACKAGE AA00025716_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


33
(Truax, Wargo, Carkhuff, TurnelI, & Glenn, 1966) to schizophrenics
(Rogers et al., 1967) to undergraduates seen at a university counseling
center (Kurtz & Grummon, 1972). The review of the broad scale applica
tions of the Barrett-Lennard Relationship Inventory should be noted in
Chapter III.
The issue has sometimes been raised that subjectivity and bias tend
to invalidate client perceptions as a valid criterion of counseling
success (Patterson, 1958, 1959; Pohlman, 1961; Rosen, 1967). One
criticism has been that unrealistic expectations may lead to inevitable
dissatisfaction with a consequent blurring of objectivity (Patterson,
1958; Severinson, 1966). Some support for this position came from a
study by Pholman and Robinson (I960) who felt their research confirmed
the existence of definite counselor preferences among clients. The study
consisted of verbal presentation of 92 counselor behaviors to a group of
psychology students who rated the behaviors in terms of degree of
annoyance or liking. Mean ratings of 73 behaviors differed signifi
cantly from the expected mean of neutrality. Furthermore, in a review
of the literature on client preferences, Rosen (1967) listed a substan
tial number of other client preferences which may influence a clients
perception and subsequent rating of counselor effectiveness.
Another variable, client preferences for specific counseling
approaches or sets of counseling techniques, may also alter the clients
ratings. Greater satisfaction was reported by Barahal, Brammer, and
Shostrom (1950) for clients counseled by "client-centered counseling"
than for clients counseled by "traditional" techniques. One must be
cautious, however, in interpreting such findings because a client's
phenomenon influenced by a variety of confounding and interacting


The investigator concluded that the HRPI predicts both client and
supervisor evaluation scores on the criterion instruments in differing
ways. Self-disclosing responses appeared to predict higher client
ratings while clarifying responses tended to predict higher supervisor
ratings in this study.
IX


34
variables. Among these variables are the clients actual problem
(Bordin, 1955), the personality of the client (Sonne & Goldman, 1957),
the degree of counselor activity (Forgy & Black, 1954), the identity of
the counselor (Grigg & Goodstein, 1957), and perhaps a gamut of other
possible variables. Patterson (1959) pointed out that one pernicious
danger in interpreting client ratings is the possible discrepancy be
tween what a client wants and feels he needs and what is in the clients
best interest. It is conceivable that a client may rate a counselor
positively when in reality the counselor may be compounding the clients
problem. In fact, some studies have suggested that clients may not be
the most objective judges of the counselor's therapeutic conditions
(Burstein & Carkhuff, 1968; Truax, 1961).
In conclusion, the preponderance of research evidence supports the
use of client perceptions and ratings of counselor effectiveness.
Nevertheless, one must be mindful of the complex variables such as
preferences, expectations, and situational factors which may impinge
upon and qualify their meaning. Therefore, a certain degree of caution
is warranted in their use; namely, they should be employed in conjunc
tion with multiple criterion measures such as supervisor ratings. In
calling for multiple criterion measures of counseling effectiveness,
Goodstein and Grigg (1959) concluded that "client satisfaction is one
important factor in any multifactor approach to the problem of effec
tiveness in counseling" (p. 23).
Summary
Both supervisor and client ratings have been researched extensively
as effectiveness criteria. Supervisor ratings have enjoyed a high level


89
explain relatively modest portions of the score variance. Secondly,
the HRPI appears to predict supervisor ratings of the counselor trainee
on the CERS with two factors at a slightly more modest level. Thirdly,
the BLRI and the CERS appear to be measuring different aspects of the
counseling arrangement, and consequently the HRPI predicts into each
criterion in different ways with different response factors. Subse
quent use of the HRPI should be made in the light of these findings,
and future investigations in turn may amplify or modify the signifi
cance of these results.


Hamrin, S. A., & Paulsen, B. P. Counseling Adolescents. Chicago:
Science Research Associates, 1950.
Hansen, J. C., Moore, G. D., & Carkhuff, R. R. The differential rela
tionships of objective and client perceptions of counseling.
Journal of Clinical Psychology, 1968, 24_, 244-246.
Hansen, J. C., & Warner, R. W., Jr. Review of research on practicum
supervision. Counselor Education and Supervision, 1971, 10,
261-272.
Hefele, T. J., & Hurst, M. W. InterpersonaI skill measurement: Pre
cision, validity and utility. The Counseling Psychologist, 1972,
3, 62-70.
Heller, K., Myer, R. A., & Kline, L. V. Interview behavior as a func
tion of standardized client roles. Journal of Consulting Psychol
ogy, 1963, 27, I 17-122.
Hill, G. E., & Green, D. A. The selection, preparation, and profession
alization of guidance and personnel workers. Review of Educational
Research, I960, 30, 115-130.
Hoffman, A. E. An analysis of counselor sub-roles. Journal of Coun
seling Psychology, 1959, 6, 61-67.
Hollenbeck, G. P. Conditions and outcomes in the student-parent rela
tionship. Journal of Consulting Psychology, 1965, ,29, 237-241.
Hough, J. B. The dogmatism factor in human relations training of pre
service teachers. Paper presented at the Educational Research
Association, Chicago, Illinois, February, 1965.
Howe, E. S., & Pope, B. The dimensionality of ratings of therapist
verbal responses. Journal of Consulting Psychology, 1961, 25,
296-300.
Hurst, M. W., & Shatkin, S. D. Relationship between standardized
admissions variables and certain interpersonal skills. Counselor
Education and Supervision, 1974, 14, 22-33.
Institute for Personality and Ability Testing, New prediction possi
bilities for vocationaI and educationaI counseI ing with the 16 PF,
Information Bulletin No. _9. Champaign, Illinois: IPAT, 1963.
Ivey, A. E. Microcounseling. Springfield, Illinois: Charles C.
Thomas, 1971.
Ivey, A. E., Normington, C. J., Miller, C. D., Morrill, W. H., & Haase,
R. F. Microcounseling and attending behavior: An approach to
pre-practicum counselor training. Journal of Counseling Psychology
Monograph, 1968, JJ5, 1-12.


122
National Vocational Guidance Association. Counselor Preparation. Wash
ington: The Association, 1949.
Nie, Norman H., Hull, C. Hadlai, Jenkins, Jean G,, Steinbrenner, Karen, &
Bent, Dale H. Statistical Package for the Social Sciences (2nd
ed.). New York: McGraw-Hill Book Company, 1975.
O'Hern, J. S., & Arbuckle, D. S. Sensitivity: A measurable concept?
Personnel and Guidance Journal, 1964, 4_2, 572-576.
Ohlsen, M. M. Final technical report on Institute for Elementary School
Counselors. Col lege of Education, University of Illinois, 1967.
Pal lone, N. J., & Grande, P. P. Counselor verbal mode, problem rele
vant communication and client rapport. Journal of Counseling
Psychology, 1965, J_2, 359-365.
Pare, D. D. Accurate Empathy Scale: Relative or Absolute? Doctoral
dissertation, University of Florida, 1970.
Patterson, C. H. A note on the evaluation of the effectiveness of
counseling and psychotherapy. Counselor Education and Supervision,
1964, 3, 129-131. (a)
Patterson, C. H. Client expectations and social conditioning. PersonneI
and Guidance Journal, 1958, 57, 136-138.
Patterson, C. H. Comments on client satisfaction, counselors, and the
counseling process. Personnel and Guidance Journal, 1959, 38, 25.
Patterson, C. H. Supervising students in the counseling practicum.
Journal of Counseling Psychology, 1964, J_l_, 47-55. (b)
Patterson, C. H. The selection of counselors. In J. M. Whiteley (Ed.),
Research in Counseling. Columbus, Ohio: Charles E. Merrill, 1967.
Peters, H. J., & Hansen, J. C. Counseling practicum: Bases for super
vision. Counselor Education and Supervision, 1963, 2, 82-85.
Pfeifle, H. H. Client perceptions of counselors with and without a
practicum. Counselor Education and Supervision, 1971, J_, 49-55.
Pierce, R., Carkhuff, R. R., & Berenson, B. G. The differential effects
of high and low functioning counselors upon counselors-in-training.
Journal of Clinical Psychology, 1967, 23, 212-215.
Pohlman, E. Changes in client preferences during counseling. PersonneI
and Guidance Journal, 1961, 40, 340-343,
Pohlman, E., & Robinson, F. P. Client reaction to some aspects of the
counseling situation. Personnel and Guidance Journal, I960, 38,
546-551.


Table 8Continued
CERS
CERS
Counsel
Supervisi on
HRPI A
HRPI B
HRPI C
BLR 1
Total
-0. 16
-0.31
0. 16
0.02
0.00
BLRI
Regard
0.02
-0. 1 1
0.25
0.01
-0.01
BLRI
Empathy
-0. 18
-0.33
0.07
-0.02
-0. 12
BLRI
Unconditiona1ity
-0.23
-0.34
0. 12
-0.06
0. 1 1
BLRI
Congruence
-0. 15
-0.27
0. 19
0. 1 1
0.05
CERS
Tota 1
0.92
0.92
0.08
0. 18
0.23
CERS
Counse1ing
1.00
0.71
0. 16
0.19
0.19
CERS
Supervision
1 .00
-0.03
0.13
0.21
HRPI
A
1.00
0.24
0.28
HRPI
B
1.00
0. 19
HRPI
C
1.00
HRPI D
HRPI E
HRP I F
HRPI G
HRPI H
HRPI I
HRPI J
HRPI D
0.06
0.00
0.05
0. I I
0.03
-0. 14
-0.22
-0.03
-0.35
0.08
-0.20
I .00
oo


Interviews: Self-initiated versus required. Journal of Counseling
Psychology, 1968, _[5, 402-406.
Brown, R. G. A. study of the perceptua I organ i zat ion of e I ementary and
secondary Outstanding Young Educators. Doctoral dissertation,
University of Florida, 1970.
Burnett, C. W. Selection and training of school and college personnel
workers. Review of Educational Research, 1954, 24, 121-133.
Burstein, J. W., & Carkhuff, R. R. Objective therapist and client
ratings of therapist-offered fcilitative conditions of moderate
to low functioning therapists. Journal of Clinical Psychology,
1968, 24, 240-241.
Cahoon, R. A. Some counseI or attitudes and characteristies re Iated to
the counseling relationship. (Doctoral dissertation, Ohio State
University) Ann Arbor, Michigan: University Microfilms, 1962.
No. 63-2480.
Cannon, J. R., & Carkhuff, R. R. Effects of rater level of function
ing and experience upon the discrimination of fcilitative condi
tions. Journal of Consulting and Clinical Psychology, 1968, 33,
189-194.
Carkhuff, R. R. Helping and human relations, volume I, selection and
training. New York: Holt, Rinehart, and Winston, Inc., 1969.
(a)
Carkhuff, R. R. Helping and human relations: A primer for lay and
professional helpers, voIume I I: Practice and research. New
York: Holt, Rinehart, and Winston, Inc., 1969. (b)
Carkhuff, R. R. Training in the counseling and therapeutic practices:
Requiem or rev¡elle. Journal of Counseling Psychology, 1966, 13,
360-367.
Carkhuff, R. R., & Alexik, M. Effect of client depth of self-explora
tion upon high- and low-functioning counselors. Journal of
Counseling Psychology, 1967, J_4, 350-355.
Carkhuff, R. R., & Berenson, B. G. Beyond counseling and therapy.
New York: Holt, Rinehart, and Winston, Inc., 1967.
Carkhuff, R. R., Collingwood, T., & Renz, L. The effects of didactic
training upon trainee level of discrimination and communication.
Journal of Clinical Psychology, 1969, 25, 460-461.
Carkhuff, R. R., Kratochvil, D., & Friel, T. Effects of professional
training: Communication and discrimination of fcilitative condi
tions. Journal of Counseling Psychology, 1968, 15, 68-74.


43
such as understanding-probing, or interpretive-evaluative to occur five
times each. The similar pairs were randomly distributed throughout the
test items. Theoretically, this method was believed to allow identifi
cation of a counselors preferred response style by quantifying the
over-chosen response categories.
Validity studies on the HRPI were conducted by Bradley (1973) in
an unpublished study of both counselor trainees and a paraprofessionaI
group of child care workers. Using the Counselor Evaluation Rating
Scale as a criterion, the regression studies for a sample of 65 counselor
trainees revealed that the HRPI scores and ratings on the CERS were
correlated .47 overall. The counseling scale of the CERS alone corre
lated .48 with HRPI scores.
In the companion investigation, HRPI scores were correlated with a
global subjective rating of child care workers by their immediate super
visors as effective, adequate, or relatively ineffective helpers. A
multiple correlation of .63 was yielded, but the understanding score
alone correlated .61 with supervisor rankings.
Test-retest reliability of the HRPI was found to be .83 for a
heterogeneous sample of 43 subjects after a span of four to six weeks
(Bradley, 1973).
Split-half reliabilities were computed by Lister (1972) for the
five original sub-scales as well as a composite scale based upon forty
consensus items agreed upon by a sample of eleven doctoral level coun
selor educators on the faculty of the University of Florida. This con
sensus scale was labeled the "expert" scale. The split-half reliabili
ties for the a priori scales and the "expert" scale were generally
modest and are summarized in Table I for three separate sample groups.


80
and supervisor evaluations on the CERS. This study also investigated
the degree of relationship among the three complete sets of variables
produced by the two criterion instruments and the predictor instrument.
A group of 46 graduate-1 eve I counselor trainees participated in
all aspects of the study along with their clients and supervisors. The
trainees completed the HRPI and were in turn evaluated by their clients
on the BLR I and by their supervisors on the CERS. By means of a step
wise regression technique the ten factor scores of the HRPI and four
selected demographic variables were regressed upon each of the criterion
instruments, employing both subscores and total scores. The findings
of this investigation were presented in Chapter IV.
Summary of the Results
Two directional hypotheses were tested during the investigation,
and each major hypothesis subsumed several hypotheses. The partial
questions break down each hypothesis into its components which are sub
scale and total scale scores within the criterion instrument. The major
hypotheses, the attendant partial hypotheses, and the results of the
analytical process on each are:
HO| There is a relationship between counselor trainee response
preferences and client evaluations of their counseling effec
tiveness. Fail to reject; partially.
1. A relationship exists between response preferences and Total
score on the BLRI. Fail to reject.
2. A relationship exists between response preferences and Level
of Regard score on the BLRI. Fail to reject.


75
demographic variable, including GPA, showed up in the set of signifi
cant predictors found in Table II. It will be noted that for the CERS
Supervision score no predictor reached the necessary .05 level of
sign!ficance.
TABLE I I
COUNSELOR EVALUATION RATING SCALE PREDICTORS
(p<.05)
Criterion/Predictors
Multiple R
R2
df
F
Total Score
Factor E (Clarification -
Superficiality)
0.2721
0.0740
1 ,44
3. 52*
Counseling Score
Factor E (Clarification -
Superficiality)
0.3208
0.1029
1 ,44
5.05*
Factor D (Analytical Interest -
Va 1idation)
0.3836
0.1472
2,43
3.71*
Supervision Score
No Predictors
*p<.05
demographic variable, including GPA, showed up in the set of significant
predictors found in Table II. It will be noted that for the CERS Super
vision score no predictor reached the necessary .05 level of signifi
cance.
The CERS Total score was predicted by only Factor E (Clarification -


16
of the counselor in the wide range of possible settings,
'~5, and
problems is obscure.
Still we have not exhausted the variables that may cc'fc
,'d an
attempt to define and measure an effective counseling rela~'c
-shi p.
Some variables that have gained attention are: the attit-Si'
a 1 and
value system of the therapist (Combs & Soper, 1963; White's"
5 Betz,
1954); demographic variables (Bailey, Warshaw, & Eichler, 195
9); the
experience of the therapist (Cartwright & Vogel, I960); the c
scree of
disturbance of the patient (Stone, Frank, Nash, & Imber, 196
; and the
clients felt need to change (Cartwright & Lerner, 1963). Ob
.iously
this list is not comprehensive, but it clearly illustrates rr
s host
of variables which have been found to influence therapeutic e
--ective-
ness. If all the variables which it is possible to study .-.er
s added to
the list of those which have been examined, the resulting cc^
scsite
would likely be astronomical.
For some the complexity of teasing effectiveness out c~
"is laby-
rinth of variables is felt to be overwhelming. A few cone _s
s _nat
counseling effectiveness research is so elusive as to ever ce
ssyond
critical inquiry (Thoresen, 1969). Some writers feel the
:-erica1
foundations of counseling have been shaken (Ford & Urban, ¡96
, and
others fear the super-structure of the profession is dangers-
s / close
to collapse into chaotic confusion and revolution (Colby, 95^
;; boltz, 1966; Rogers, 1963). Simultaneous with this expresses
concern
is the growing pressure for the profession to be increasi nc :
account
able. The ground swell of public demand calls for the course
's pro-
fession to demonstrate selection, evaluation, and education c-
'icedures
which are derived from and justified by a substantial body e~
e-~s i ri ca 1
research on effectiveness (Thoresen, 1969).


45
oblique factor matrix was clear with ten factors accounting for 60.2%
of the total score variance and 87.6% of the common variance. Factors
were limited to items that had at least a .30 factor loading. Factor
scores were obtained then by weighting each item in proportion to its
factor loading as follows:
Factor Scores
Factor Loading Range Weight of the Item
.30 to .39 I
.40 to .49 2
.50 to .59 3
.60 to .69 4
.70 to .79 5
The relative independence of the oblique factor structure is
attested to by the low correlations between primary factors. All inter
correlations are below .30 except between factors A and J, which is
0.38.
The 48 items, which were the variables factor analyzed, presented
alternative choices of responses perceived as being helpful. In each
item a value judgment was required by the respondent of which item to
choose and which to reject. Consequently, every response represented
a preference of one response over another one. The factors corres
pondingly have a high end which represents the responses chosen and a
low end which represents the responses rejected. The factor labels,
which are believed to describe the common theme of the items loading in
each factor group, are listed below. Also presented are the number of
items loading the factor and the total weights for the factor. Factors
are listed in descending order according to the proportion of total
score variance accounted for by the factor.


CHAPTER I I I
METHODOLOGY
This study attempted to investigate the extent to which effective
ness ratings of counselor trainees were related to preferred communica-
tional modes. Preferred verbal responses of counselor trainees were
examined in relationship to both client perceptions of counseling rela
tionships and supervisory evaluations of counseling effectiveness.
Description of the Sample
The subjects involved in this study included University of Florida
counselor trainees, the trainees' clients, and supervisors of the
trainees. A more detailed description of these samples, including the
process of selection, follows.
I. Counselor Trainees. The counselors-in-training group in this study
were composed of graduate students in the Department of Counselor
Education at the University of Florida. These counselor trainees
were all currently enrolled in a practical, supervised counseling
experience, referred to as a practicum or an internship, through the
Department of Counselor Education. Prior to this supervised train
ing period each trainee had completed an introductory course in the
helping process followed by a course which combined a survey of
counseling theories and approaches with a counseling laboratory.
The laboratory experience utilized role playing and video-taped
microcounseling (Ivey, 1971) with volunteer clients. Both of these
36


83
human pattern. The presence of significant observed differences may be
more remarkable when considered in this light.
Interpretation and Discussion of Findings
Although the two major hypotheses could not be accepted in full,
there were certain discernible trends apparent in the findings. Clearly
the HRPI factors are predicting into two different spheres represented
by client perceptions of the relationship on one hand and supervisor
evaluations of the counselors ability on the other. Six HRPI factors
achieved the .05 confidence level of significance as predictors. Four
predicted the BLRI scores and two predicted the CERS scores, but no
single factor predicted into both instruments.
The client perceptions registered on the BLRI appeared to place con
siderable value upon self disclosure. Frank Feedback was the single most
potent predictor in this study for any individual criterion score, pre
dicting the Level of Regard score. The same score also was predicted
in a lesser amount by responses from an external frame of reference which
in a sense supports the self disclosure theme. Along this same line the
BLRI Total score was predicted in part by premature judgments or quick
interpretations. The implication seems to be that clients perceive
honest, direct, and even potentially threatening feedback as concerned
involvement. In fact, in this study they experienced premature judg
ments, as opposed to focused probing, as a higher level of empathy.
It is worth noting that the frank honesty valued by clients did
not appear to be unbridled bluntness. For instance, empathic under
standing also was a significant predictor of Level of Regard. Despite
the potential threat of frank feedback from an external point of


26
confronts the he I pee and himself is directly related to the
degree to which the helpee is able to confront himself and
others.
Corollary III. The degree to which the helper both acts and
directs the actions of the helpee immediately in the present
to the relationship between helper and helpee is related to
the helpee's ability to act with immediacy and later to
direct the actions of others.
Corollary IV. The degree to which the helper can make con
crete a course of constructive action is related to the
degree to which the helpee can go on to make concrete courses
of action for himself and others, (pp. 84-90)
The complete text of Carkhuff's working assumptions emphasizes
the direct personal effect upon the helpee of the helper's level
of facilitative conditions offered in the interpersonal encounter.
It follows that the degree to which the counselor in the helping role
provides high levels of empathy, warmth, respect, genuineness, and
concreteness is directly related to the internalization of these
facilitative dimensions into his own life. Moreover, the degree to
which the helper is openly and deeply himself, self-disclosing, con-
frontive of himself and his client, present with immediacy, and active
with clear, concrete, action-oriented alternatives is again directly
related to the client's capacity to translate these same facilitative
dimensions into his own life sphere.
Carkhuff has written extensively about the selection process and
its relationship to the previously stated propositions. He has also
addressed himself to the effectiveness criteria problems already pre
sented. His position is that "the best index of a future criterion is
a previous index of that criterion" (1969b, p. 85). In general, the
counselor selection process must be distilled to identification of
specific functioning within the helper role by the prospective helper.


123
Polmantier, P. C. The personality of the counselor. Vocational Guid
ance Quarterly, 1966, 16, 95-100.
Poole, A. Counselor judgment and counselees evaluation. Journal of
Counseling Psychology, 1957, 4_, 37-40.
Porter, E. H. Therapeutic Counseling. Boston: Houghton Mifflin Com
pany, 1950.
Purcell, F. E. Counseling assignments and efficiency. VocationaI
Guidance Quarterly, 1957, 5_, MI-113.
Reed, M. A study of client satisfaction as a criterion to evaluate
effective counseling. Unpublished master's thesis, University of
Utah, 1969.
Rapaport, A. The promise and pitfalls of information theory. In Buck-
ley, W. (Ed.), Modern Systems Research for the Behavioral Scien
tist. Chicago: Aldine, 1968.
Redfering, D. L. Changes in perception of the counselor: Self, ideal
self, and self as judged by peers. Counselor Education and Super
vision. 1973, _[3, 289-293.
Rickabaugh, K., Heaps, R. A., & Finley, R. Counselor comfort, counseling
climate, and client satisfaction: Client ratings and academic
improvement. Counselor Education and Supervision, 1972, JJ_, 219-
223.
Riese, H. C., & Stoner, W. G. The meaning of words: Teachers and
counselors. Counselor Education and Supervision, 1973, 13, 46-51.
Rogers, C. R. A tentative scale for the measurement of process psycho
therapy. In E. A. Rubinstein and M. B. Parloff (Eds.), Research
in Psychotherapy, Vo Iume I. Washington, D. C.: American Psycho
logical Association, 1959, 96-107.
Rogers, C. R. A process conception of psychotherapy. The American
Psychologist, 1958, J_3, 142-149.
Rogers, C. R. Psychotherapy todayor, where do we go from here?
American Journal of Psychotherapy, 1963, 17, 5-16.
Rogers, C. R. The necessary and sufficient conditions of therapeutic
personality change. Journal of Counseling Psychology, 1957, 21,
95-103.
Rogers, C. R., & Dymond, R. F. (Eds.). Psychotherapy and Personality
Change. Chicago: University of Chicago Press, 1954.
Rogers, C. R. Gendlin, E. T., Kiesler, D. J., & Truax, C. B. The Thera
peutic Relationship and its Impact: A_ Study of Psychotherapy with
Schizophrenics. Madison: University of Wisconsin Press, 1967.


95
43. He approves of some things I do, and plainly disapproves of
others.
44. He is w¡I Iing to express whatever is actually in his mind with
me, including any feelings about himself or about me.
45. He doesnt like me for myself,
46. At times he thinks that I feel a lot more strongly about a
particular thing than I really do,
47. Whether I am in good spirits or feeling upset does not make
him feel any more or less appreciative of me.
48. He is openly himself in our relationship.
49. I seem to irritate and bother him.
50. He does not realize how sensitive I am about some of the
things we discuss.
51. Whether the ideas and feelings I express are "good" or "bad"
seems to make no difference to his feeling toward me.
52. There are times when I feel that his outward response to me
is quite different from the way he feels underneath.
53. At times he feels contempt for me.
54. He understands me.
55. Sometimes I am more worthwhile in his eyes than I am at other
times.
56. I have not felt that he tries to hide anything from himself
that he feels with me.
57. He is truly interested in me.
58. His response to me is usually so fixed and automatic that I
dont really get through to him.
59. I don't think that anything I say or do really changes the
way he feels toward me.
60. What he says to me often gives a wrong impression of his
whole thought or feeling at the time.
61. He feels deep affection for me.
62. When I am hurt or upset he can recognize my feelings exactly,
without becoming upset himself.
63. What other people think of me does (or would, if he knew)
affect the way he feels toward me.
64. I believe that he has feelings he does not tell me about that
are causing difficulty in our relationship.


ACKNOWLEDGMENTS
The author wishes to express sincere appreciation to the following
people who assisted in this study:
Dr. Joseph Wittmer, the chairman of the committee, whose leader
ship, experience, and precious time were invaluable assets.
Dr. Larry C. Loesch whose expertise in research design and statis
tical analysis was essential to this study and deeply appreciated.
Dr. Paul G. Schauble whose sensitivity, encouragement, and insight
were major contributions.
Dr. James L. Lister whose instrument provided the primary focus of
this study and who originally served as the writer's committee chairman.
Ms Barbara Rucker whose skills contributed both to the computer
analysis and to the splendid job of typing this study.
Barbara, my wife, whose love, patience, and understanding were
primary sources of inner strength during my entire graduate school
experience. It is to her that this study is appreciatively and affec
tionately dedicated.
ui


Ded¡cated to
Barbara Sherrill Bradley


54
performance in a supervised practicum (Myrick & Kelly, 1971). The items
are designed to assess the counselor trainee's understanding of a
counseling rationale, counseling practice with counselees, and explora
tion of self and counseling relationships with a supervisor. A 7-point
(+3 to -3) Ukert-type scale is employed to mark each item. The CERS
yields three scores: a counseling subscale (C) is an appraisal of the
counselor-trainee's individual counseling behavior; a supervision sub
scale (S) is an assessment of the person's involvement and progress in
the supervisory process; and a total score (T) of these two subscales
plus one additional item: "Can be recommended for a counseling posi
tion without reservation." The resulting total score ranges between 27
and 189 when the negative scores are eliminated. The counseling scale
and the supervision scale each yield scores between 13 and 91.
In their original article (1971) the authors, Myrick and Kelly,
reported a reliability study involving 45 student counselors and their
supervisors from the University of Florida. Individual supervisees,
who met with their assigned faculty supervisor for a minimum of one
hour per week for twelve v/eeks, were rated by these supervisors on the
CERS at the end of a fall quarter practicum. Utilizing the Spearman-
Brown correction of the split-half reliability procedure, the investi
gators found a coefficient of .95. When the 13 counseling items were
compared with the 13 supervisor items, a correlation coefficient of .86
was produced.
The stability of the CERS over a period of time was determined by
a test-retest reliability procedure. After a minimum of a four week
lapse in time since collecting the CERS ratings, the supervisors were
again requested to rate the same student counselors with whom they had


37
courses lasted for an academic quarter and preceded the first super
vised practicum. Students pursuing a specialist degree must com
plete three practica which typically amount to ten hours per week
and one internship quarter which is defined as a full-time, forty
hours per week counseling experience. For doctoral students the
total requirement is five practica and three quarters of intern
ship. Counselor trainees in this study were at any point in the
completion of their degree programs from the first practica to the
last internship. Trainees have available actual counseling oppor
tunities in a variety of academic and community settings. Counselor
trainees in this study were, for example, involved in a university
counseling center, a university mental health service, a community
mental health center, a drug treatment center, a delinquency treat
ment program, a paraprofessionaI training program, a community
college, a public school, a hospital, or any of a number of similar
human service agencies either affiliated with the University of
Florida or in nearby communities.
Graduate counseling students who were involved exclusively with
elementary and middle school children were not included in this
study since children that young often have difficulty completing the
inventories used in this study. All other counselor trainees
involved in a practicum or internship during the target quarter,
and who received supervision through the department, were asked to
participate in the data collection process.
2. Supervisors. The supervisors in this study were counselor educators
or professional counselors who had themselves completed rigorous
graduate training programs in counseling at the university level.


APPENDIX A
COUNSELOR TRAINEE DEMOGRAPHIC DATA SHEET
Your name:
Phone number:
Sex:
Age:
Undergraduate major:
Cumulative graduate grade point average at the end of last quarter:
Number of graduate quarters you will have completed at the end of this
quarter:
Number of quarters of supervised counseling practice (practicum or
internship) you will have completed at the end of this quarter:
Individual supervisor:
Present practicum or internship agency:
91


19
outcome than those who saw their counselors as passive listeners. In
his 1961 report Brams emphasized the strong tie between the counselor's
personality and his ability to communicate with clients, and Pal lone
and Grande (1965) even concluded that the counselor's verbal mode sig
nificantly influences the amount of the client's probIem-reIevant
communication. Clearly the counselor's communication style is funda
mental to the relationship established with the client. However, the
extent of this influence has not been established. This study, in part,
will be an attempt to evaluate the impact of verbal response preference
to a multi-factor measure of counseling effectiveness.
In the belief that the counselor's communicational mode is a direct
expression of one's unique perceptual world, and that in turn one's
communicational style crucially shapes the atmosphere of the counseling
relationship; the focus of this review now turns to the last two
approaches to counseling effectiveness relevant to this study: the
perceptual approach and the core therapeutic conditions approach as
epitomized in the writings of Carkhuff.
The Perceptual Model
A much different approach to the question of counselor effective
ness has been introduced by perceptual psychology. According to this
approach (Combs & Snygg, 1959), the internal perceptions and perceptual
organization of the individual provide a meaningful perspective for the
examination of effectiveness in human helping. The assumption is made
that counselors who have "learned to use themselves as effective instru
ments in the production of helping relationships can be distinguished
from those who are ineffective on the basis of their characteristic


121
MacGuffie, R. A., Janzen, F, V., & McPhee, W. M. The expression and
perception of feelings between student and supervisors in a practi-
cum setting. Counselor Education and Supervision, 1970, _9, 263-266.
McClain, E. W. Sixteen Personality Factor Questionnaire scores and
success in counseling. Journal of Counseling Psychology, 1968,
J_5, 492-496.
McDougall, W., & Reitan, H. The use of a peer rating technique in
appraising selected attributes of counselor trainees. Counselor
Education and Supervision, 1961, J_, 72-74.
Mason, J., & Blumberg, A. Perceived educational value of the classroom
and teacher-pupil interpersonal relationships. Journal of Secondary
Education, 1969, 44, 135-139.
McDaniel, W. Counselor selection: An evaluation of instruments. Coun
selor Education and Supervision, 1967, 6_, 142-144.
Mcllvaine, J. F. Coached clients as raters of counseling effectiveness.
Counselor Education and Supervision, 1972, 12, 123-129.
McNair, D. M., Callahan, D. M., & Lorr, M. Therapist "type" and patient
response to psychotherapy. Journal of Consulting Psychology, 1962,
26, 425-429.
McWhirter, J. J., & Marks, S. E. An investigation of the relationship
between the facilitative conditions and peer and group leader
ratings of perceived counseling effectiveness. JournaI of Clinical
Psychology, 1972, 28, I 16-117.
Mendoza, B. F. Predicting Counselor Effectiveness: A Multiple Regression
Approach. Doctoral dissertation, Western Michigan University,
1966.
Mills, D. H., & Zytowski, D. G. Helping relationship: A structural
analysis. Journal of Counseling Psychology, 1967, 14, 193-197.
Mueller, W. J., Gatsch, C. M., & Ralston, J. K. The prediction of
counselor interview behavior. Personnel and Guidance Journal,
1963, £l_> 513-517.
Munger, P. F., & Johnson, C. A. Changes in attitude associated with an
NDEA counseling and guidance institute. Personnel and Guidance
Journal, I960, 38, 751-753.
Myrick, R. D., & Kelly, F. D. A scale for evaluating practicum students
in counseling and supervision. Counselor Education and Supervision,
1971, J_0, 330-336.
Myrick, R. D., Kelly, F. D., and Wittmer, J. The Sixteen Personality
Factor Questionnaire as a predictor of counselor effectiveness.
Counselor Education and Supervision, 1972, I 1, 293-301.


TABLE 8
INTERCORRELATION MATRIX FOR ALL VARIABLES OF THE BARRETT-LENNARD RELATIONSHIP
INVENTORY, THE COUNSELOR EVALUATION RATING SCALE, AND
THE HELPER RESPONSE PREFERENCE INVENTORY
BLRI
BLRI
BLRI
BLRI
BLRI
CERS
Tota 1
Regard
Empathy
Unconditiona1ity
Congruence
Tota 1
BLRI Total 1.00
0.84
0.93
0.67
0.90
-0.26
BLRI Regard
1.00
0.74
0.47
0.73
-0.05
BLRI Empathy
1.00
0.47
0.84
-0.28
BLR 1 Unconditiona1ity
1 .00
0.40
-0.31
BLRI Congruence
1.00
-0.23
CERS Tota 1
1 .00
CERS Counsel¡ng
CERS Supervision
HRPI A
HRPI B
HRP I C
HRP I D
HRP I E
HRP I F
HRP I G
HRP I H
HRP I I
HRP I J
CTi


17
Perhaps the most viable alternative is for the profession to move
more toward the "systems approach" to conceptualizing and evaluating
counseling. A key concept is that the system is more than the sum of
its parts. For example, the counseling encounter Is a total relation
ship built upon two distinct but temporally engaged personalities
attempting to bridge their separateness via a communications process.
The crucial variables may be the unique perceptual worlds of the coun
selor and client, the communication that flows between them, and the
emotional tone of the atmosphere within which they transact. Perhaps
this triad offers a viable and comprehensive approach to the enormously
complex give-and-take of the counseling process. Since the variables
under investigation in this study relate specifically to communication,
it is important to consider the relevance of this factor to counseling.
The Communications Analysis Model
Fiedler (1950, 1951) pointed out the crucial significance of the
relationship to counseling. He reported that experts excel non-experts
in their ability to communicate and understand their patients. Porter
(1950) went a step further and hypothesized that the counselor's style
of preferred responses sets the tone of the interview and defines the
counselor's capacity to be helpful. In Porter's instruments, the Porter
Test of Counselor Attitudes and the Porter Interview Analysis Scale, he
categorized responses as understanding, supportive, i nterpretive,
probing and evaluative. A number of studies of trainees in counselor
education programs have revealed that with exposure to training and
supervision trainees become more understanding and less evaluative,
supportive, and probing (Demos & Zuwaylif, 1963; Kassera & Sease, 1970;


105
Helper A:
"Youre angry because 1 had to punish you yesterday!"
Helper B:
"Please sit down. Were going to have a nice surprise!"
22. He 1 pee:
"Why does he do that to me? 1 can't understand how he
can keep it up?"
Helper A:
"You really resent it, don't you?"
Helper B:
"Maybe he really doesn't intend to hurt you."
23. Helpee:
"I've never tried to smoke pot, but most of my friends
do. 1 realize it's just a matter of time before 1
check it out."
Helper A:
"Of course, you know the risks you would be taking if
you give in to their pressures."
Helper B:
"What do you expect to get out of it, when you try it?"
24. Helpee:
"Mrs. Brown, I've finished my math assignment, but Sally
thinks 1 haven't done it right. Can 1 look it over
again and turn it in tomorrow?"
Helper A:
"No, Mary. 1 want you to give it to me now. You know
it's important for you to do your own work."
Helper B:
"I'd like you to turn it in now, Mary. Let me look it
over and try not to worry about the mistakes."
25. Helpee:
"1 think 1 want to sign up again for dramatics. You
know, 'cause that's the one thing in school so far that
really turns me on."
He 1 per A:
"This is one thing you really like."
Helper B:
"Maybe you like dramatics because it gives you more of
a chance to express yourself."
26. Helpee:
"Last night 1 had this dream in which 1 was a drill
sergeant, and 1 was getting a big kick out of shoving
the recruits around."
Helper A:
"Maybe you have a greater need for power than youve
realized up to now."
Helper B:
"Can you think of some real life situations where you
get that same satisfaction from being in charge?"
27. Helpee:
"1 came from a very happy home. In fact, there was
never a quarrel or disagreement between my parents, but
my wife and 1 seem to be fighting constantly."


119
Jackson, D. D. The study of the family: Family rules. Family Process,
1965, 12, 589-594.
Jennings, G. D. The relationship between perceptual characteristics
and effective advisinq of university housing para-professionaI
residence assistants. Doctoral dissertation, University of Florida,
1973.
Johnston, J. A. Practicum and on-the-job ratings of school counselors.
Personnel and Guidance Journa I, I 966, _45, 16-19.
Jones, A. J. Principles of Guidance and Personnel Work. New York:
McGraw-HiII, 1951.
Jones, L. K. The counselor evaluation rating scale: A valid criterion
of counselor effectiveness? Counselor Education and Supervision,
1974, J_4, I 12-1 16.
Kagan, N., & Krathwohl, D. (Eds.). Studies in Human Interaction: Inter
personal Process RecaI I St i mu Iated by V ideotape. U. S. Department
of Health, Education, and Welfare Project, No. 5-0800, December,
1967.
Kassera, W. J., & Sease, W. A. Personal change as a concomitant of
counselor education. Counselor Education and Supervision, 1970,
9, 208-21 I.
Kaul, T. J., Kaul, M. A., & Bednar, R. L. Counselor confrontation and
client depth of self-exploration. Journal of Counseling Psychology,
1973, 20, 132-136.
Keltz, J. W. The development and evaluation of a measure of counselor
effectiveness. Personnel and Guidance Journal, 1966, 44, 511-516.
Kemp, C. G. Influence of dogmatism on the training of counselors.
Journal of Counseling Psychology, 1962, 9_, 155-157.
Kemp, D. E. Correlates of the Whitehorn-Betz AB scale in a quasi-
therapeutic situation. Journal of Consulting Psychology, 1966,
30, 509-516.
Kerlinger, F. N. Foundations of Behavioral Research. New York: Holt
Rinehart and Winston, 1973.
Kiesler, D. J. Basic methodological issues implicit in psychotherapy
process research. American Journal of Psychotherapy, 1966, 20,
135-155.
Kilpatrick, F. P., & CantriI, H. Self-anchoring scaling: A measure of
individuals' unique reality worlds. Journal of Individual Psy
chology, I960, J_6, 158-174.


49
TABLE 3
TEST-RETEST RELIABILITY OF THE SUBTESTS OF THE
64-ITEM RELATIONSHIP INVENTORY
IN SEVERAL STUDIES
Relationship
Dimensions
Barrett-Lennard
(1969)
Form OS,
N=40
2 to 6 weeks
Barrett-Lennard
(1969)
Form MO,
N=38
12 days
Mills and
Zytowski
(1967)
Form MO,
N=79
3 weeks
Mills and
Zytowski
(1967)
Form OS,
N=79
3 weeks
Level of
Regard
.88
.79
.86
.74
Empathic
Understanding
.86
.91
.84
.90
Unconditiona-
lity of Regard
.86
.86
.80
.80
Congruence
.92
.85
.87
.88
Tota 1 Score
.92
.89
It must be recognized that evidence supporting the validity of the
Barrett-Lennard Relationship Inventory is cumulative rather than con
clusive. The bulk of evidence derives from research, much of which was
carefully designed, in which the associations between Relationship
Inventory measures and other variables were predicted on the basis of
logical and theoretical grounds. When these findings are positive, the
underlying theory gains credibility; and the argument that the dimen
sions, purported to be measured, are indeed related to measurable rela
tionship variables of fundamental importance in the consequences of
human interaction becomes more believable. The assumption, then, was
that a mounting body of findings which were consistent with the constructs


20
perceptual organizations"(Combs & Snygg, 1959, p. 14), The perceptual
approach then may be contrasted with more behavioristic approaches which
operationalize the criteria of effective helping in terms of specifiable
behaviors.
According to the perceptual approach, a man's behavior is a func
tion of his "perceptual field," By perceptual field is understood the
individual's unique perceptions of himself and the world in which he
lives. Hence, the field consists of what the person perceives at a
given point in time and the meaning or significance he attaches to that
perceptual experience (Combs, Avila, & Purkey, 1971). Behavior is seen,
therefore, as a direct consequence of the individual's field of meanings
existing for him at the moment of behaving. Since the perceptual
psychologist is concerned with internal impression rather than external
expressions, the internal frame of reference is preeminent. Perhaps it
would not be an overstatement to say that the perceptual theorist
believes that a counselor is effective not because of his specific
behaviors but rather because of the internal perceptions behind those
behaviors.
Walton and Sweeney (1969) cite the study of counselor perceptions
by Combs and Soper (1963) as one of the more promising approaches to the
measurement of effectiveness. Using a critical incidents in human
relationships format, they found moderately high positive relationships
between 12 characteristic ways of perceiving and effectiveness as a
counselor. The effective counselor was sensitive to and concerned with
how things looked to others and was oriented to people rather than
things. He perceived others as able rather than unable, dependable
rather than undependable, friendly rather than unfriendly, and worthy


48
of the Relationship Inventory were acceptable and generally consistent
across investigations. Barrett-Lennard (1962) reported split-half
reliability coefficients for the various relationship dimensions within
a range of .82 to .93. In two separate studies, Snelbecker (1961,
1967) discovered a range of split-half reliabilities between .75 and
.94. The range was .83 to .95 for Hollenbeck (1965) and .82 to .91
for Hough (1965).
The test-retest reliability coefficients reported by Barrett-
Lennard (1962) after a four-week interval, ranged from .84 to .90. After
a six-months interim between administrations Hollenbeck (1965) found a
somewhat lower range of reliability, on the 72-item form, of from .61
to .81. In an investigation using the group form of the Relationship
Inventory, Berzon (1964) reported a reliability coefficient of .86 for
total scores after a four-week interval.
Refinement procedures have helped the reliability coefficients for
the 64-item revision of the Relationship Inventory to run as high or
higher than with previous versions of the instrument. Several test-
retest studies are summarized in Table 3.
The question of the validity of the Barrett-Lennard Relationship
Inventory is interwoven with the complexities of obtaining direct cri
terion-based validity checks in the overall field of counseling.
Barrett-Lennard (1969) relates this difficulty to "the absence of alter
native, established measures of theoretically equivalent dimensions of
perceived interpersonal response" (p. 4). Although a number of
researchers have been basically in agreement with the statement by Mills
and Zytowski (1967) that "the test has, at a minimum, face validity,"
(p. 194) a more persuasive justification for its use as a criterion
instrument is essential.


I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
P. Jo^£>h WiTtmer, Chairman
Professor of Counselor Education
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Paul G. Schauble
Professor of Psychology
This dissertation was submitted to the Graduate Faculty of the College
of Education and to the Graduate Council, and was accepted as partial
fulfillment of the requirements for the degree of Doctor of Philosophy.
December, 1976
Dean?Graduate School


29
(Eberlein & Park, 1971), and the Counselor Evaluation Rating Scale
(Myrick & Kelly, 1971), which was one of two criterion measures employed
in this research study.
There have been a variety of other unnamed rating scales including
a sorting process (Embree, 1954), and free response type counselor
perception instrument (Arbuckle & Wicas, 1957), sociometric indices
(Gade, 1967; Steffi re, King, & Leafgren, 1962), a categorized rating
instrument (Keltz, 1966), a sensitivity scale (O'Hern & Arbuckle, 1964),
and a number of semantic differential rating scales (Bartlett & Thomp
son, 1971; Redfering, 1973; Riese & Stoner, 1973; and Shapiro, 1968).
It can be said that rating scales have served a large number of
researchers as an accessible, simplified, and quantifiable approach to
the complex problem of finding a viable, valid effectiveness criterion.
In general, the above instruments were validated on the basis of accept
able correlations with other measures of counseling effectiveness such
as rankings or "expert" judgments. At times validity has been inferred
from reliability studies. While such practices are defended by respected
researchers in the field (Combs & Soper, 1963; Russo, Kelz, & Hudson,
1964; Steffi re, King, & Leafgren, 1962), it should be noted that actual
client outcome criteria were not generally used in the validation of the
above instrument.
Although the use of rating scales has extended into many areas that
are not germaine to this study, it can be noted that considerable
research has been generated using counselor self ratings, peer ratings,
coached client ratings, independent judge ratings, and a wide spectrum
of combinations. Nevertheless, this study focused specifically upon
the use of supervisor and client ratings as the basis for effectiveness


in people, patient, sensitive to others, emotionally stable, objective,
respectful of facts, and trusted by others.
9
Hamrin and Paulsen in 1950 attempted to introduce a slightly more
scientific approach. They conducted a field study of 91 counselors to
arrive at a consenses of desirable traits which they believed facilitated
counseling. In descending order of frequency they listed I) under
standing, 2) sympathetic attitude, 3) friendliness, 4) sense of humor,
5) stability, 6) patience, 7) objectivity, 8) sincerity, 9) tact, 10)
fairness, II) tolerance, 12) neatness, 13) calmness, 14) broadminded
ness, 15) kindliness, 16) pleasantness, 17) social intelligence, and
18) poise.
By I960, however, Hill and Green reported that the enthusiastic
search for the essential personality pattern of a counselor had all but
dwindled away. The focus by then had shifted from speculative lists of
character traits to examining the total personality of the counselor
and how that personality impacts upon the counseling relationship. This
transition naturally arose out of a growing disenchantment with the
subjectivity of the philosophicaI-specuI at ive approach (Cottle, 1953;
Hill and Green, I960; Jones, 1951). While no massive revolution in the
understanding of effective counseling resulted from the philosophical
pursuit of ideal counselor characteristics, such outgrowths as Rogers'
formulation (1957) concerning empathy, genuiness, and unconditional
regard proved to be of considerable heuristic value. Cottle (1953)
recognized the helpfulness of the trait listings but criticized them
because the lists (I) were merely opinions, (2) failed to distinguish
counselors from related professionals, (3) varied widely, and (4)
failed to specify which patterns or interreI at ions of traits were


6
pursuing either a specialist or doctoral degree at the University
of Florida. Each trainee was currently engaged in either a super
vised practicum or internship in counseling at the time of data
coI Iection.
3. Client Ratinqs The perception by a client of the helpful quality
of the relationship with the counselor trainee. Operationally, the
client rated the counselor trainee on the Barret-Lennard Relation
ship Inventory. Higher scores were assumed to be indicating of a
more helpful relationship in the perception of the client.
4. Supervisor Evaluations The perception of the counselor trainee's
effectiveness by his individual supervisor. The ratings were
operationalized as scores on the Counselor Evaluation Rating Scale.
Higher scores indicated the supervisor perceived the counselor
trainee to be relatively more effective as a helping person.
5. Counseling Effectiveness The abiIity to a counselor trainee to
beneficially affect a client. For the purposes of this study,
counseling effectiveness was operationally defined in terms of
ratings of the counselor-in-training by clients and supervisors.


74
primary predictor and only other predictor, the GPA. Also in each case
Factor F was a negative predictor.
The single strongest predictor in this study related to the Level
of Regard score. For that score Factor J (Frank Feedback Incongruence)
accounted for 17$ of the score variance. GPA added another 8$ to the
prediction process. Factor A (Empathic Understanding Probing) and
Factor I (Internal External Frame of Reference) each explained 4$
of the variance. As a result the multiple correlation rose to .57, the
highest in this research, and collectively the four predictors accounted
for 32$ of the Level of Regard score variance.
In addition to the 11$ of variance attributed to the GPA in the
Unconditionality of Regard score Factor J (Frank Feedback Incongruence)
again entered the equation in the amount of 6$ of variance accountability.
However, it should be noted that in this case, the variable which was
the strongest positive predictor of Level of Regard became a negative
predictor of Unconditionality of Regard.
Although the results are generally modest, ostensibly somewhat
contradictory, and involve only four of the ten HRPI factors, and only
four of the BLRI scores, there does appear to be a significant, obser
vable trend between client ratings and response preference scores on the
HRPI. Therefore, with due regard to the need for further clarifica
tion, the directional hypothesis HO| is accepted for the purposes of
this study.
Test of the Second Major Hypothesis
Even though the four demographic variables were also used in the
regression of the second criterion, supervisor ratings on the CERS, no


103
Helper B:
"What kinds of things seem to lead up to the feelings
youre describing now?"
10. Helpee:
"Ive been a teacher for seventeen years, and Ive
always put the kids first, and look where that's got
me!"
Helper A:
"But you really have been right, you know. You really
should put the growth of the children first."
Helper B:
"You seem to be rationalizing your current situation by
saying that you haven't been appreciated for being a
good teacher."
II. He 1 pee:
"My parents are pretty determined that 1 should stay at
home and attend the local junior college. 1 feel that
I'm being headstrong and rebellious, but 1 just cant
live at home any longer, college or no college!"
Helper A:
"It's probably best for you to get out, since you feel
so strongly about it."
Helper B:
"What seems to be the trouble at home? Are your parents
not getting along with each other?"
12. Helpee:
"If the boss leans on me one more time like he did
today, I'm going to give it to him straight. I've just
about had it!"
Helper A:
"1 know how you feel, but you'll probably feel a lot
better after you've blown off a little of the steam to
me."
Helper B:
"You feel pretty sore at him, don't you?"
13. Helpee:
"1 don't see why 1 can't do it. All of my friends are."
Helper A:
"But, 1 don't want you to feel bad about it."
Helper B:
"You seem to want to justify your behavior through the
actions of others."
14. Helpee:
"No matter how hard 1 try, 1 still don't feel that 1
measure up or that people feel that I'm good enough for
them."
Helper A:
"But others may not feel that way about you."
Helper B:
"It's not good for you to let others influence you so
much."
15. Helpee:
"Why should 1 study harder? No matter how well prepared
1 am, 1 still don't have a fighting chance against the
frat guys who always have copies of the exams."


APPENDIX B
COVER LETTER TO THE CLIENT
To the cIient:
Your counselor has been asked to find one client who will rate him/her
as a counselor. You are the only client selected by your counselor to
take part in this research so your participation is extremely important.
You should know that how you rate your counselor will not be used to
grade or evaluate him/her personally. Rather all of the ratings will
be combined and examined as a group. Nevertheless, it is very impor
tant that we receive one hundred per cent cooperation. The name of
your counselor appears on your rating sheet only for the purpose of
identifying which rating scales have not been returned.
You will notice that the rating scale is stamped and addressed to make
your task as easy as possible. I realize it will take some time to
complete the Inventory, but your efforts will be sincerely appreciated.
Please read the instructions carefully before you begin.
Thank you,
Sidney L. Bradley
Doctoral Candidate
Department of Counselor Education
92


51
that "open" teachers would see themselves as responsible for diffi
culties and remedial action in their classroom work. The rationale was
that such teachers would tend to be more genuine, sensitive, and
accepting than teachers who, under similar conditions, see problems as
unrelated to their own characteristies and externally imposed. The
finding that students of the more "open" teachers did perceive them more
positively on each of the Relationship Inventory dimensions may be
interpreted as implying that the measures discriminated differences
consistent with theoretical predictions.
Cahoon (1962), in another well-known investigation, found signifi
cant relationships between the process experiential levels and open-
mindedness, on one hand, and the client-perceived quality of the coun
seling relationships on the other. For his sample of practicum-Ievel
counselor trainees he employed the Process Scale, the Dogmatism Scale,
and the Barrett-Lennard Relationship Inventory.
The uses of the Relationship Inventory have been extensive.
Donnan, Harlan, and Thompson (1969) used the original 85-item form to
significantly correlate four personality factors on the 16 PF with
three relationship variables for a sample of 22 counselors who were
evaluated by 880 prospective college freshmen following counseling. In
another study, Brown and Calis (1968) measured the effect upon the
clients perception of the relationship with his or her group counselor,
depending upon whether the group was required or self-initiated. Lanning
(1971) used the Relationship Inventory to evaluate whether practicum
students perceive the supervisory relationship the same in group and
individual supervision. A still different application of the Inventory
was made by Armstrong (1969) who examined the difference between students


21
rather than unworthy. Moreover, he perceived himself as being identi
fied with people rather than apart from people, as personally adequate
rather than wanting, and as seif-revealing rather than self-concealing.
These dichotomous characteristics which Combs and Soper studied
grew out of a graduate seminar held in 1959 at the University of Florida
(Combs, 1961). The purpose of the seminar was to investigate effective
helping relationships from the internal point of view. The effort to
identify perceptual characteristics which were believed to be relevant
to the task of effective professional helping crystallized in five focal
areas (I) the general frame of reference from which the helper approaches
the situation, (2) the ways in which the helper perceived other people,
(3) the ways in which the helper perceived himself, (4) the ways in which
the helper perceived the task with which he was confronted, and (5)
the ways in which the helper perceived appropriate methods for carrying
out his purposes. The forty dichotomies issuing from that seminar are
listed below. Those perceptual dichotomies believed related to effec
tive helpers are on the left and those related to ineffective helpers on
the right.
(I) General Frame of Reference
InternaI
Growth orientation
Perceptual meanings
PeopIe
HopefuI
Causation oriented
Externa I
Fencing in or controlling
Facts, events
Things
Despairing
Mechanics oriented
(2) Perceptions of Other People Sees others as:
Capable
T rustworthy
HelpfuI
Unthreatening
Respectable
Worthy
Incapab I e
Untrustworthy
Hindering
Threatening
No account
Unworthy


47
unconditionality of regard. The original instrument contained 85 items
but was soon shortened to 72 items, tost of the research on the Rela
tionship Inventory before mid-1964 utilized the 72-item "interim revi
sion" of the original investigative instrument. Since that time, the
64-item revision, or adaptations of this revision, has been employed
in most research using the Inventory. The author (1969) explains the
procedures used in revising the instrument to its current form:
The 64-item revision was based (a) on item-analysis results
from several samples of data obtained with earlier versions
of the instrument,(b) on other technical considerations,
such as achieving a numerical balance of positively and
negatively stated items in each of the four scales, (c) on
minor theoretical refinements, particularly in respect to
the unconditionality of regard scale and (d) on the con
cern to alter or replace items of a relatively abstract
or difficult kind. The item analysis provided empirical
checks on what each previously used item was doing and
gave results that were used in conjunction with b), c)
and d) to refine the selection and preparation of items.
In general, the aim of the revision was to further im
prove the sensitivity and versatility of the instrument
for measuring the defined variables in the context of any
significant interpersonal relationship. (Barrett-Lennard,
p. I, 1969)
Well over one hundred research studies have used the Barrett-
Lennard Relationship Inventory establishing the instrument as one
of the most widely used measures of the quality of human relation
ship available today. In its current form, the Relationship Inven
tory's 64 items include 16 for each of the four subtest dimensions,
which assess a person's perception of another person in a relation
ship. Each statement is scored on a 6-point scale, from +3, "yes,
I strongly feel that is not true." There is no 0 response. The
items are general enough to assess practically any two-person rela
tionship which the examiner may specify.
The published reliability studies on the earlier 72-item form


57
In the report on validity studies using the Helper Response Prefer
ence Inventory, it was pointed out that for a sample of 65 counselor

trainees there was a .47 correlation between HRPI scores and overall
CERS scores. The correlation was .48 for the counseling scale of the
CERS when related to the HRPI scores separately.
Borman and Ramirez (1975) compared the CERS self ratings of 25
master's degree counseling students in practicum with their ratings by
their supervisor and practicum assistants who were doctoral students.
Their findings revealed that there were significant differences found on
9 of the 27 items. However, most of the differences were attributed to
the students rating themselves higher than supervisors or assistants
and, specifically, differences in ratings between students and practicum
assistants. On only one item, dealing with the degree of relaxation and
comfort in the supervisory sessions, was there a statistically signifi
cant difference between the supervisor's ratings and the counselors'
self-ratings.
Analysis of Data
The data collection process, previously described, yielded four
basic groupings of data: demographic variables, Helper Response Pre
ference Inventory scores, Barrett-Lennard Relationship Inventory scores,
and Counselor Evaluation Rating Scale scores. Using the two criterion
instruments eight dependent criterion variables were isolated: the
Relationship Inventory total score plus the four subscores, and the
counseling, supervisory, and total scales from the Counselor Evaluation
Rating Scale. By means of the SPSS library computer program for stepwise
regression (1973), the ten factor scores of the Helper Response


44
TABLE I
SPLIT-HALF RELIABILITIES FOR THREE SAMPLES: GRADUATE LEVEL COUNSELOR
TRAINEES, PARAPROFESSIONAL COUNSELOR TRAINEES IN A COMMUNITY COLLEGE,
AND PARAPROFESSIONAL DISASTER RELIEF COUNSELOR TRAINEES
Santa Fe Community
College Human Service
Aide Trainees
(N=6I)
Counselor Education
Graduate Students
(N-72)
Wilkes-Barre Flood
ParaprofessionaI
Emergency Counselor
T rainees
(N=50)
Expert
.62
.62
.49
Understanding
.53
.52
.52
Supportive
.41
.47
.42
1nterpretive
. 17
.42
.34
Probing
.18
.30
.52
Evaluative
.13
.37
.29
Factor analysis of the HRPI was conducted in 1973 (Bradley) pooling
319 sets of responses in the sample. Since the items were forced
dichotomies, it was imperative to compute tetrachoric correlations. The
items were factor analyzed, and the principal axis solution rotated to
varimax criterion (Guertin & Bailey, 1970). Highest coefficients
served as communality estimates. Two items had to be deleted because
of low cell distributions. The remaining 48 item variables were factor
analyzed and again rotated to varimax criterion, but a clear factor
structure did not emerge until the principal axes factor matrices were
rotated in an oblique rotation procedure. Ten principal axes factors
were chosen because the latent roots tended to plateau beyond that point
and all ten factors chosen had latent roots greater than 1.0. The


109
Helper B:
"1 know its been hard for you, but 1 dont feel that
you've ever resorted to deliberate deception.
45. Helpee:
"Ive been engaged for six months now, and my fiance
wants to live together, even though we wont be able to
get married until we both graduate. I'm really in con
flict over what to do."
Helper A:
"Is this partly because you are afraid of what he'll
think of you if you dont give in?"
Helper B:
"It hasn't been an easy thing to decide, has it?"
46. Helpee:
"1 wish 1 could have more self confidence. Everyone 1
know seems to have it."
Helper A:
"What are the things you don't feel confident about?"
Helper B:
"You feel this is something missing that you really
want."
47. Helpee:
"Just because 1 failed to promote six kids last year,
that doesn't mean I'm not a good teacher. But 1 suppose
that's just not the way to get along around here."
Helper A:
"Are you confident that the kids you failed really
weren't making the grade?"
Helper B:
"Could you be using the failures as the reason for the
negative feedback you're getting?"
48. Helpee:
"Do you think I'll ever be able to pass this course?
I've just looked over the texts and the reading list,
and I'm afraid that it's just too much for me."
Helper A:
"1 suspect that most students have that same reaction at
the beginning of the course. But you realize that most
of them do actually pass the course, and some do very
we 11."
Helper B:
"1 wonder why you feel this way. Have you had some bad
experiences with this subject in high school?"


65
TABLE 6
MEANS, STANDARD DEVIATIONS, AND RANGES OF
SCORES ON THE COUNSELOR EVALUATION
RATING SCALE (N=46)
CERS
Sea 1es
Means
Standard
Deviations
Ranqe
Counseling
rO
fO
oo
10.35
33-91
Supervision
84.22
9.98
38-91
Total
171.54
20.05
91-189
Supervision, are comparable. Both mean scores fall slightly less than
one standard deviation below the maximum possible score, a phenomenon
observed as well for the 171.54 mean of the Total scores.
Scores of the Counselor Trainees on the Predictor Instrument
The means, standard deviations, and ranges of scores on the ten
factor scales of the Helper Response Preference Inventory are presented
in Table 7. It should be kept in mind that each factor represents a
chosen direction of responses and a rejected direction of responses.
Higher scores reflect a preference for the response group named first in
the parenthetical pair, and lower scores indicate that the response
group placed second in the parenthesis was preferred. All of the means
fall in the upper half of possible score range for the factor except
for Factor D (Analytical Interest Validation) where the preference
favored slightly the validation side of the dichotomous scale. For
four factors the entire score range observed fell in or very close to


(3) Perceptions of Self Sees self as:
Identified with people
Enough
T rustworthy
Li ked
Wanted
Accepted
Feels certain, sure
Feels aware
Self-reveaIing
(4) The Helping Task and Its Prob
Purpose is:
Helping
Larger meanings
Altruistic
Understanding
Accepting
Valuing Integrity
Approach to problem is:
Positive
Open to experience
Process oriented
Relaxed
Awareness of complexity
Tolerant of ambiguity
(3) Perceptions of Methods Sees
Apart from people
Wanting
Not trustworthy
Not Iiked
Not wanted
Not accepted
Doubt
Unaware
Se If-conceaIing
ems
Dominating
SmaI Ier meanings
Narcissistic
Condemning
Rejecting
Violating integrity
Negative
Closed to experience
Ends oriented
Compulsion to change others
Overs impI ification
Intolerant of ambiguity
appropriate methods as:
Helping
Cooperation
Asserting
Acceptance
Permissive
Open communication
Giving
Vital
Man ipuI ating
Competition
Appeasing
Rejecting (attacking)
Authoritarian
Closed communication
Withholding
Lifeless
(Combs, 1961, pp. 56-57)
More than twenty studies have attempted to infer the perceptual
organization of helpers using this variety of original perceptual dimen
sions plus a few that have been identified subsequently. Among the
more notable studies are those of priests (Benton, 1964), secondary
school teachers (Brown, 1970), counselor trainees (Combs & Soper, 1963)


116
Driekurs, R., & Sonstegard, M. A specific approach to practicum super
vision. Counselor Education and Supervision, 1966, 6^ 18-26.
Dunlop, R. S. Counselor competence: Some proposals in search of
advocacy. Personnel and Guidance Journal, 1968, _46, 655-660.
Ebel, R. L. Estimation of the reliability of ratings. Psychometrika,
1951, 16, 407-424.
Eberlein, L., & Park, J. Se If-concept/idea I-self-concept congruence
and rated effectiveness of counselor trainees. Counselor Educa
tion and Supervi sion, 1971, J_0, 26- 132.
Embree, R. B. The assessment of counselor competence through the use
of multiple ratings by supervisors and fellow students. American
Psycho I oqi st, 1954, 9_, 360.
Emmerling, F. C. A study of the relationship between personality
characteristies of classroom teachers and pup i I perceptions of
these teachers. (Doctoral dissertation, Auburn University) Ann
Arbor, Michigan: University Microfilms, 1961.
FauIkenberry, P. D. A study of the evaluation of counseling effective
ness based upon supervision, peer, role pIayer, and self-ratinqs.
(Doctoral dissertation, Purdue University) Ann Arbor, Mich.:
University Microfilms, 1968. No. 68-12, 550.
Fiedler, F. E. Factor analysis of psychoanalytic, nondirective, and
adlerian therapeutic relationships. Journal of Consulting
Psychology, 1951, J_5, 32-38.
Fiedler, F. E. The concept of an ideal therapeutic relationship.
Journal of Consulting Psychology, 1950, J_4, 239-245.
Flanagan, J. C. The critical incident techniques. Psychological
Bulletin. 1954, 51, 327-358.
Foley, W. J., & Proff, E. C. NDEA institute trainees and vocational
rehabilitation counselor: A comparison of characteristics.
Counselor Education and Supervision, 1965, 1, 154-159.
Ford, D. H., & Urban, H. B. Psychotherapy. In Farnsworth, P. R.,
(Ed.) Annual Review of Psychology, Vol. 13, Palo Alto, California:
Annual Reviews, 1967.
Forgy, W. E., & Black, J. D. A follow-up after three years of clients
counseled by two methods. Journal of Counseling Psychology, 1954,
1, 1-8.
Form, A. L. The construction of a scale on attitudes toward counseling.
Journal of Counseling Psychology, 1955, 7_, 96-102.
Forster, J. R., & Flamburg, R. L. Further exploration of the 16 PF


LIST OF TABLES
Table Page
1 Split-Half Reliabilities for Three Samples: Graduate Level
Counselor Trainees, Paraprofess ionaI Counselor Trainees
in a Community College, and ParaprofessionaI Disaster
Relief Counselor Trainees 44
2 Factor Labels and Factor Scores for the Helper Response
Preference Inventory 46
3 Test-Retest Reliability of the Subtests of the 64-1 tern
Relationship Inventory in Several Studies 49
4 Demographic Data on the Participating Counselor Trainee
Subjects 62
5 Means, Standard Deviations, and Ranges of Scores on the
Barrett-Lennard Relationship Inventory (N=46) 64
6 Means, Standard Deviations, and Ranges of Scores on the
Counselor Evaluation Rating Scale (N=46) 65
7 Means, Standard Deviations, and Ranges of Scores on the
Helper Response Preference Inventory (N=46) 66
8 I ntercorrelat ion Matrix of All Variables of the Barrett-
Lennard Relationship Inventory, the Counselor Evaluation
Rating Scale, and the Helper Response Preference
Inventory 67
9 Mean Correlation of Each Helper Response Preference
Inventory Factor with All of the Scores of the Two
Criterion Instruments 71
10 Barrett-Lennard Relationship Inventory Predictors 73
11 Counselor Evaluation Rating Scale Predictors 75
12 Regression Equation for the Barrett-Lennard Relationship
Inventory Predictors 77
13 Regression Equation for the Counselor Evaluation Rating
Scale Predictors 78


107
Helper B:
"1 really doubt it. With your good scores in general
aptitude and reading, you should do well in almost
everything besides math."
33. Helpee:
"Its finally coming clear to me. 1 had always believed
that people were usually disappointed in me, but now it
doesn't seem that that's been the case."
Helper A:
"So, now you're having some different ideas about how
people react to you."
Helper B:
"It must be a good feeling to find that people haven't
been so disappointed in you all along."
34. Helpee:
"My principal doesn't seem to understand what I'm trying
to do in this course. I'm afraid he won't recommend my
appointment for next year."
Helper A:
"He shouldn't drop you just because he doesn't under
stand what you're doing."
Helper B:
"But this is early in the year. I'm sure he'll feel
differently long before the contracts are issued."
35. Helpee:
"My friends really don't understand me. We have a lot
of laughs together, but they don't really know what I'm
1ike."
Helper A:
"Gets kinda lonely at times?"
Helper B:
"What are the things about you that you think your
friends don't know?"
36. Helpee:
"I'm afraid to go home anymore. It seems like there's
always an argument with someone yelling or screaming at
you."
Helper A:
"This must be pretty unpleasant for you."
Helper B:
"Is this because people in your home aren't able to
really communicate with one another?"
37. Helpee:
"The psychologist who gave me a 11 of those tests would
n't tell me what the results were. Could you give me
some indication of how 1 did?"
Helper A:
"Were there some tests in particular that you were
worried about?"
Helper B:
"We may be able to discuss your results, but, you know,
these aren't the kinds of tests you pass or fail."


Blocher, D. H. A multiple regression approach to predicting success
in a counselor education program. Counselor Education and Super
vision, 1963, 3, 19-22.
Blumberg, A. Supervisory behavior and interpersonaI relations. Educa
tional Administration Quarterly, Spring, 1968, 34-45.
Board, F. A. Patients and physicians' judgments of outcome of psycho
therapy in an outpatient clinic. Archives of General Psychiatry,
1959, i, 185-196.
Bordin, E. S. Implications of client expectation for the counseling
process. Journal of Counseling Psychology, 1955, 2, 17-21.
Blocker, D. H. A multiple regression approach to predicting success in
a counselor education program. Counselor Education and Super
vision, 1963, 3, 19-22.
Bordin, E. S. Implications of client expectation for the counseling
process. Journal of Counseling Psychology, 1955, 2^, 17-21.
Bordin, E. S., Cutler, R. L., Dittmann, A. T., Harway, N. L., Raush,
H. L., & Rigler, D. Measurement problems in process research in
psychotherapy. Journal of Consulting Psychology, 1954, 18,
79-85.
Borman, C. A., & Ramirez, C. Evaluating counseling practicum students.
Counselor Education and Supervision, 1975, 15, 48-54.
Bowler, E. M., & Dawson, F. T. Counseling Employees. New York:
Prentice-Hall, 1948.
Bradley, F. 0. A modified interpersonal process recall technique as a
training model. Counselor Education and Supervision, 14, 34-39.
Bradley, S. L. Examination of the Helper Response Preference Inven
tory. Unpublished mimeograph report, University of Florida, 1973.
Bradley, S. L. Factor analysis of the Helper Response Preference
Inventory. Unpublished mimeograph report, University of Florida,
1973.
Brammer, L. M., & Shostrom, E. L. Therapeutic Psychology. Englewood
Cliffs, N. J.: Prentice-Hall, Inc., I960.
Brams, J. M. Counselor characteristics and effective communication in
counseling. Journal of Counseling Psychology, 1961, 25-30.
Brown, D., & Cannaday, M. Counselor, counselee, and supervisor ratings
of counselor effectiveness. Counselor Education and Supervision,
1969, 8, I 13-118.
Brown, 0. B., & Cala, V. F. Two methods of initiating student


39
The focus of supervision was to review the trainees counseling efforts,
primarily via tape recordings and verbal report; to assess the trainee's
level of skill; and to guide the supervisees development toward full
professional competence. The method used to evaluate the effectiveness
of the supervisee was the Counselor Evaluation Rating Scale (see Appen
dix) which the supervisor returned to the researcher following comple
tion.
During the first week of data collection, the investigator arranged
with all relevant group supervisors to meet with their groups for the
purpose of administering the Helper Response Preference Inventory (see
Appendix) to each counselor trainee. At that time demographic data on
each trainee were also gathered by means of a questionnaire (see Appen
dix). The first two demographic variables, sex and major field, were
included to obtain some descriptive data on the counselor trainee sample.
They were omitted from the statistical analysis since the data were not
continuous. Altogether the following six sets of demographic informa
tion were requested:
1. Sex of the counselor trainee;
2. Undergraduate field of major;
3. Age of the counselor trainee;
4. Cumulative graduate grade point average at the end of the
previous quarters;
5. Number of graduate quarters completed at the end of the quarter
of data col lection;
6. Number of quarters of supervised counseling practice completed
at the end of the data collection quarter.
At each group supervision session the counselor trainees were


86
selection, training, or skill analysis. These questions will be con
sidered separately in light of the empirical findings.
The nature of a concurrent, cross-sectional investigation such as
this does not directly address the issue of counselor trainee selection.
To properly evaluate an instruments usefulness as a screening tool in
the candidate selection process a longitudinal study of the instru
ments predictive validity is necessitated. Nevertheless in the real
world of practical affairs those persons charged with the responsi
bility of candidate selection must base those decisions upon the best
available information. This instrument does provide some useful
information despite the modest variances accounted for, particularly
when used in conjunction with other indicators.
Perhaps the first question that must be resolved relates to which
definition of counseling effectiveness one wants to predict. Clearly
the HRPI predicts in different ways into both client and supervisor
definitions of effectiveness. Maybe the question is not an either-or
proposition since both criteria appear to be predictable from a single
test administration. Actually a candidate could be considered in the
light of his scores on the four factors which predict client ratings and
the two which predict supervisor ratings. One must do so, neverthe
less, with full awareness that thus far the instrument has not been
investigated in a manner thoroughly consistent with such use.
Whether the HRPI should be used for training or skill analysis is
another matter. While the predictive power is much too low to make life
changing decisions on this basis, a counselor trainee may gain some
insight into his own communication style through the instrument. For
example, if the counselor trainee's profile reveals a preference away


126
Truax, C. B. Therapist empathy, warmth, and genuineness and patient
personality change in group psychotherapy: A comparison between
interaction unit measures, time sample measures, and patient
perception measures. Journal of Clinical Psychology, 1966, 22,
225-229.
Truax, C. B. The relationship between the patient's perceptions of the
level of therapeutic conditions offered in psychotherapy and con
structive personality change. Brief Research Reports, Wisconsin
Psychiatric Institute, University of Wisconsin, 1962. (b)
Truax, C, B., & Carkhuff, R. R. Concreteness: A neglected variable in
research in psychotherapy. Journal of Clinical Psychology, 1964,
20, 264-267.
Truax, C. B., & Carkhuff, R. R. Experimental manipulation of thera
peutic conditions. Journal of Consulting Psychology, 1965, 29,
119-124.
Truax, C. B., & Carkhuff, R. R. Toward Effective Counseling and Psycho
therapy. Chicago: Aldine Publishing Co., 1967.
Truax, C. B., & Mitchell, K. M. Research on certain therapist inter
personal skills in relation to process and outcome. In A. E.
Bergin and S. L. Garfield (Eds.) Handbook of psychotherapy and
behavior change: An empirical analysis. New York: John Wiley and
Sons, 1971.
Truax, C. B., Silber, L. D., & Wargo, D. G. Personality Changes and
Achievement in Therapeutic Training. Arkansas Rehabilitation
Research and Training Center, University of Arkansas, 1966.
Truax, C. B., Wargo, D. G., Carkhuff, R. R., & Tunnel, B. T., Jr.
Client perception of therapist empathy, warmth, and genuineness
and therapeutic outcome in group counseling with juvenile delin
quents. Unpublished manuscript, Arkansas Rehabilitation Research
and Training Center, University of Arkansas, 1966.
Truax, C. B., Wargo, D. G., Frank, J. D., Imber, S. D., Butle, B. C.,
Hoehn-Sarie, R., Nash, C. H., & Son, A. R. Therapist empathy,
genuineness, and warmth and patient therapeutic outcome. JournaI
of Consulting Psychology, 1966, 50, 395-401.
Truax, C. B., Wargo, D. G., Tunnel, B. T., Jr., & Glenn, A. W. Patient
perception of therapist empathy, warmth, and genuineness and thera
peutic outcome in outpatient group therapy. Unpublished manu
script, Arkansas Rehabilitation Research and Training Center,
University of Arkansas, 1966.
Van Der Veen, F. Dimensions of client and therapist behavior in rela
tion to outcome. Proceedinqs of the 73rd annua I convention of the
American Psychological Association. Washington, D. C.: American
Psychological Association, 1965, 279-280.


CHAPTER IV
RESULTS OF THE STUDY
Introduction
The research procedures for this study were enacted according to
the experimental design described in Chapter III. Analysis of the data
was conducted via the computer facilities of the Northeast Regional Data
Center under the auspices of the Florida State University System. Spe
cifically, the stepwise regression subprogram and the canonical corre
lation subprogram of the Statistical Package for the Social Sciences
developed by Nie et al. (1975) were the analytical procedures utilized.
The results of these statistical operations are presented below in a
tabular format. This chapter consists of: (I) a descriptive profile
of the counselor trainee sample, including demographic data on the
subjects, and the scores of all subjects on the two criterion instru
ments as well as the predictor instrument; and (2) the statistical tests
of the two major hypotheses and eight minor hypotheses subsumed under
the two major ones, presented in sequential order.
Descriptive Data of the Sample
The Counselor Trainees
There were 67 counselor trainees who appeared to meet the criteria
for inclusion: namely, having clients on a continuing basis who
possessed the literacy skills requisite to valid completion of the
60


and counselor effectiveness. Counselor Education and Supervision,
1976, J_5, 184-188.
Freedman, S. A., Antenen, W. W., & Lister, J. L. Counselor behavior
and personality characteristics. Counselor Education and Super-
vi sion, 1967, 26-30.
Friesen, D. D., & Dunning, G. B. Peer evaluation and practicum super
vision. Counselor Education and Supervision, 1973, 12, 229-235.
Froehlich, C. P. Preparation of teachers and specialists for guidance
service. Review of Educational Research, 1951, 21, 159-166.
Gade, E. M. The relationship of sociometric indices and counselor
candidate effectiveness. Counselor Education and Supervision, 1967,
6, 121-124.
Garrison, M. A correlation study of factors utilized in counselor
selection. Unpublished specialist project, Western Michigan
University, 1967.
Gooding, C. T. An observational analysis of the perceptual organiza
tion of effective teachers. Doctoral dissertation, University of
Florida, 1974.
Goodstein, L. D., & Grigg, A. F. Client satisfaction, counselors and
the counseling process. Personnel and Guidance Journal, 1959, 38,
19-24.
Graves, J. T. The employee counselor. Occupations, 1944, 22, 495-497.
Grigg, A. E. Client response to counselors at different levels of
experience. Journal of Counseling Psychology, 1961, 8^ 217-223.
Grigg, A. E., & Goodstein, D. D. The use of clients as judges of the
counselor's performance. Journal of Counselinq Psychology. 1957,
4, 31-36.
Gross, W. F., Curtain, M. E., & Moore, K. B. Appraisal of a milieu
therapy environment by treatment team and patients. Journal of
Clinical Psychology, 1971, 26, 541-545.
Gross, W. F., & DeRidder, L. M. Significant movement in comparatively
short-term counseling. Journal of Counseling Psychology, 1966,
13, 98-99.
Guertin, W. H., & Bailey, J. P. Introduction to Factor Analysis. Ann
Arbor: Edwards Brothers, 1970.
Haley, J. Strategies of Psychotherapy. New York: Grue and Stratton,
1963.


63
Supervisors
The supervisors in this study numbered 27. There were 20 males and
7 females. With the exception of one psychiatrist all of the super
visors held an earned doctorate in counseling, guidance, student personne
work, or some allied branch of psychology. Two supervisors gave
individual supervision to four trainees, and therefore each rated four
counselor trainees. Three supervisors completed ratings on the three
trainees they supervised individually, and seven supervisors rated two
trainees. The remaining fifteen supervisors each contributed a single
set of ratings on one counselor trainee. Only two female supervisors
completed more than one trainee's ratings, and in both cases the number
of ratings was two.
Clients
To intentionally preserve a sense of anonymity there was no attempt
to gather descriptive information on the client sample. While such knowl
edge would of course be interesting, there would be no way to compare
the clients who responded with those who for one reason or another never
returned the Relationship Inventory. Consequently, client data would be
of limited interpretive value.
Client Evaluation Scores on the Counselor Trainees
Converted scores from the evaluative raw data are presented in
composite form in the following two tables. The client ratings on the
Barrett-Lennard Relationship Inventory are collapsed into means,
standard deviations, and ranges in Table 5. The scores are character
ized by broad ranges with 46 to 47 points of spread among the four sub
scales. The total score range is 182 points.


13
Zuwaylif, 1966), the Self-Description Checklist (Wicas & Mahan, 1966),
the California Personality Inventory and the MMPI (Schroeder & Dowse,
1968), and the Interpersonal Check List (Eberlein & Park, 1971). While
some of these studies have reported promising avenues for reliable
prediction, frequently the attempts to replicate have either yielded
conflictual findings or qualified the original results in ways that
diminish confidence in the ultimate dependability of the instrument.
An excellent illustration of this frustrating process has been the
attempts to differentiate the measurable characteristics of A and B
type therapists as first described by Whitehorn and Betz (1954, I960).
They found that type A therapists who established warm, personal rela
tionships had substantially better success rates with schizophrenics
than the more passively detached and interpretive type B therapists.
Furthermore, they found that 23 items on the Strong Vocational Interest
Blank distinguished the two therapist types. However, McNair, Callahan,
and Lorr (1962) found that with neurotic patients, and psychologists as
therapists, the results were reversed and the type B therapists were
more successful. Carson et al. (1964) concluded from their research
that the type A therapist works best with distrusting clients and type
B with trusting clients, in an attempt to reconcile the discrepancies.
Kemp (1966) introduced still more controversy by concluding from his
research that both types of therapists are most capable of helping those
clients who produce the most discomfort in themselves. It is readily
apparent that the findings are intriguing but decidedly inconclusive.
The Criteria Crisis in Effectiveness Model Building
The disappointing attempts to link measurable personality variables


27
He suggests the following specific indices: I) an index of communica
tion; 2) an index of trainabiIity; 3) an index of self exploration; and
4) a critical incident index to ascertain his potential helpfulness in
the face of crisis. This study relates to the first and last of these
indices because the predictor instrument is a measure of communication
within a criticaI-incidents or counseling-simulation format.
Summary
Despite voluminous research with a myriad of approaches, the effec
tive counselor remains a nebulous concept. The speculative-philosophi
cal approach helped clarify for training purposes the ideal characteris
tics and proved thought-provoking for a period, but the approach came
under critical fire because its subjectivity did not solidify the base
of the profession.
Empiricism, in the form of psychometric testing, came into vogue
as attention focused upon the measurable personality traits of the
counselor. There was not found, however, many impressive links between
specific personality traits and beneficial client outcome. It appeared
that a more global assessment of the total impact of the helpers
personality rather than a trait-factor approach was more relevant to the
therapeutic involvement of a truly helpful relationship.
Perceptual dimensions have attracted considerable attention in the
counseling profession, but research on this approach has been less than
prolific. Focus upon the internal frame of reference of the counselor
has yielded some meaningful links with effective helping.
More recently, the pressures within and without the profession
toward accountability have led to attempts to define counseling


APPENDIX E
COUNSELOR EVALUATION RATING SCALE
Counselor Date
Supervisor
Below are listed some statements which are related to evaluation in
supervising a counseling experience. Please consider each statement
with reference to your knowledge of the counselor rated.
Mark each statement in the left hand blank according to how strongly
you agree or disagree. Please mark every statement. Write in +3, +2,
or +1, or -I, -2, or -3, to represent the following:
+3:
1 strongly agree
-1 :
1 s1ight1 y disagree
+2:
1 agree
-2:
1 disagree
+ 1 :
1 s1ight1y agree
-3:
1 strongly disagree
1. Demonstrates an interest in client's problems.
2. Tends to approach clients in a mechanical, perfunctory manner.
3. Lacks sensitivity to dynamics of self in supervisory relation-
shi p.
4. Seeks and considers professional opinion of supervisors and
other counselors when the need arises.
5. Tends to talk more than client during counseling.
6. Is sensitive to dynamics of self in counseling relationships.
7. Cannot accept constructive criticism.
8. Is genuinely relaxed and comfortable in the counseling session.
9. Is aware of both content and feeling in counseling sessions.
10. Keeps appointments on time and completes supervisory assign
ments.
11. Can deal with content and feeling during supervision.
12. Tends to be rigid in counseling behavior.
13. Lectures and moralizes in counseling.
14. Can critique counseling tapes and gain insights with minimum
help from supervisor.
15. Is genuinely relaxed and comfortable in the supervisory session.
99


72
Testing the Hypotheses
Test of the First Major Hypothesis
The forward stepwise multiple regression technique was applied to
all scores of the two criterion instruments, the BLRI and the CERS,
using the ten factor scores of the HRPI as predictor variables. Four
demographic variables (age, grade point average, number of graduate
quarters completed, and number of quarters of supervised counseling
practice completed) were also entered into the computations as possible
predictors.
The first hypothesis postulated that higher client ratings could
be predicted from counselor trainee response preferences as measured
by the HRPI, the predictor instrument. Table 10 presents the predictors
that met the .05 level of significance required for inclusion in the
regression equation for each of the five BLRI scores. The strongest
predictor overall was graduate grade point average (GPA). In fact GPA
was the best predictor of the Total score in addition to three of the
subscores. GPA was the only variable which significantly predicted the
Congruence subscale. The GPA accounted for 8$ of the score variance in
the Level of Regard, the Empathy Level, and the Congruence scores. The
variance accounted for by the GPA was II# of the Unconditionality of
Regard score and rose to 13# of the Total score. No other demographic
variable was a significant predictor of any of the BLRI scores.
HRPI scores were significant predictors on four of the five BLRI
scores. Factor F (Focused Probing Premature Judgments) accounted for
5# of the variance in both the Total score and the Empathy Level score.
In both cases Factor F explained a lesser amount of variance than the


46
TABLE 2
FACTOR LABELS AND FACTOR SCORES FOR THE HELPER
RESPONSE PREFERENCE INVENTORY
Preferred Response
Factor Label
Rejected Response
Factor Label
Number
of Items
Total (
Weight
A.
Empathic Under
stand i ng
Probing
12
27
B.
Confrontation
False Reassurance
9
21
C.
Specificity
Ambiguity
8
20
D.
Va 1idation of
Experience
Analytical Interest
6
12
E.
Clarificat ion
Superficia1ity
9
16
F.
Focused Probing
Premature Judgment
5
1 1
G.
Tentative Formu-
1 at ions
Avoidance of Emotion
4
7
H.
Requesting In
formation
Giving Information
4
8
1.
Helpee's Intern
al Reference
Helpers External
Reference
5
8
J.
Frank Feedback
1 ncongruence
5
10
Barrett-Lennard Relationship Inventory (BLRI)
The theoretical rationale which led to the construction of the
Barrett-Lennard Relationship Inventory (Barrett-Lennard, 1962) was
derived from the formulation by Carl Rogers (1957) of "necessary and
sufficient conditions" for beneficial client change. The Relationship
Inventory was intended to assess four aspects of any two-person rela
tionship, particularly the counselor-client relationship. Those four
aspects were empathic understanding, level of regard, congruence, and


5
counseling-simulation instrument have a significant relation
ship to clients' perceptions of a beneficial relationship?
2. Does a counselor trainee's response preferences on a printed
counsel ing-simuI at ion instrument have a significant relation
ship to supervisors' evaluations of counseling effectiveness?
Hypotheses
The general hypothesis in this research was that a relationship
would be found between the response preferences of counselor trainees
on the experimental instrument and two separate measures of counseling
effectiveness; namely, supervisor ratings and client perceptions of the
relationship. The following sub-hypotheses were tested:
HO| There is a relationship between counselor trainee response
preferences and client evaluations of their counseling
effectiveness.
H02 There is a relationship between counselor trainee response
preferences and supervisor evaluations of their counseling
effectiveness.
Definition of Terms
In this study the following definitions of relevant terms apply:
1. Counselor Response Preference The choice of a verbal reply to a
client's stimulus statement. In this study the helpee statement on
each item of the Helper Response Preference Inventory was the
stimulus. The counselor trainee was asked to choose one of two
helper responses for his reply. Since the format was forced-choice
the range of options was limited to only two response possibilities
2. Counselor Trainee Graduate students in Counselor Education


129
Stephen. Currently he is Director of the Child and Family Enrichment
Center and Vice President of the Advent Christian Home. He is a member
of the National Association of Social Workers and the Academy of Certi
fied Social Workers. He also holds clinical membership in the American
Association of Marriage and Family Counselors. Mr. Bradley holds
membership in the following honorary organizations: Delta Epsilon Chi,
Phi Kappa Phi, and Kappa Delta Pi.


APPENDIX C
RELATIONSHIP INVENTORY-FORM OS-M-64
Below are listed a variety of ways that one person may feel or behave
in relation to another person.
Please consider each statement with reference to your present relation
ship with your
Mark each statement in the left margin, according to how strongly you
feel that it is true, or not true, in this relationship. Please mark
every one. Write in +3, +2, +1, or -I, -2, -3, to stand for the
following answers:
+3:
Yes, 1 strongly feel that it
-1 : No,
1 feel that it is probably
is true.
untrue or more untrue than true.
+2:
Yes, 1 feel it is true.
-2: No,
1 feel it is not true.
+ 1 :
Yes, 1 feel that it is
-3: No,
1 strongly feel that it is
probably true or more true
than untrue.
not
true.
1. He respects me as a person.
2. He wants to understand how I see things.
3. His interest in me depends on the things I say or do.
4. He is comfortable and at ease in our relationship.
5. He feels a true I iking for me.
6. He may understand my words but he does not see the way I feel.
7. Whether I am feeling happy or unhappy with myself makes no
real difference to the way he feels about me.
8. I feel that he puts oh a role or front with me.
9. He is impatient with me.
10. He nearly always knows exactly what I mean.
11. Depending on my behavior, he has a better opinion of me some
times than he has at other times.
12. I feel that he is real and genuine with me.
13. 1 feel appreciated by him.
14. He looks at what I do from his own point of view.
15. His feeling toward me doesnt depend on how I feel toward him.
16. It makes him uneasy when I ask or talk about certain things.
93


62
TABLE 4
DEMOGRAPHIC DATA ON THE 46 PARTICIPATING
COUNSELOR TRAINEE SUBJECTS
Variable
Age
Mean
27.57
Standard deviation
5.24
Median
26.00
Range
22-51
Sex
Ma 1 e
27
(58.7$)
Female
19
(41.3$)
Grade Point Average
Mean
3.77
Standard deviation
.20
Range
3.20-4.00
Number of Graduate Quarters Completed
Mean
7.63
Standard deviation
3.26
Range
3-17
Number of Quarters of Supervised
Counseling Comp
1 eted
Mean
4. 10
Standard deviation
2.26
Range
1-8
Undergraduate Major Field
Psychology
27
(58.7$)
Education
5
(10.9$)
Language
5
(10.9$)
Sociology
2
( 4.3$)
Physical Sciences
2
( 4.3$)
Re 1 igi on
2
( 4.3$)
Political Science
1
( 2.2$)
Art History
1
( 2.2$)
Business
1
( 2.2$)


87
from frank feedback or clarification, he may be alerted to ways he can
enhance his effectiveness in terms of his clients perception, his super
visors perception, or both. Likewise, he may discover that his candor
is not balanced by understanding or vice versa, sensitizing himself to
areas worthy of emphasis in his ski I I-bui Iding process. In short, the
Helper Response Preference Inventory shows promise, albeit tentative,
as a useful educational tool to help bridge the gap between self dis
closure and empathic understanding and between the effectiveness models
of clients and supervisors.
Suggestions for Further Research
An examination of the findings of this study suggests that further
research in a number of areas may be beneficial. In general the ration
ale that gave rise to this study and the literature review in Chapter II
speak clearly of the need for development of instruments designed to
measure counseling effectiveness. Specifically there is a need for
instruments to assess the communication style of relatively effective
counselors in a continuing effort to tease out the elements distinctively
characteristic of helpful communication.
Previous references have been made to the need for a longitudinal
study to evaluate the effectiveness over time of counselor trainees who
have taken the h'RPI. The goal would be to arrive at a predictive formula
whereby a probability statement could be made regarding a counselor
trainee who attains a given score on the instrument. Of course, such
precision is a long way off and will doubtless prove to be elusive if
not unattainable. At the very least a thorough examination of how a
trainees HRPI score profile changes over time, especially with
practical supervised training, could be highly instructive.


BIBLIOGRAPHY
Abeles, N. A study of the characteristics of counselor trainees. Pis-
sertation Abstracts, 1958, 18, 2204-2205.
Abramowitz, S. I., & Jackson, C. The comparative effectiveness of
there-and-then versus here-and-now therapist interpretations in
group psychotherapy. Journal of Counseling Psychology, 1974, 21,
288-293.
ACES Committee on Standards for Counselor Education Secondary.
Standard for the preparation of secondary school counselors 1967
edition. Memorandum to ACES members, January 0, 1967.
Allen, T. W. Effectiveness of counselor trainees as a function of
psychological openness. Journal of Counseling Psychology, 1967,
14, 35-40.
American Personnel and Guidance Association. A statement of policy.
The counselor: Professional preparation and role. Personnel and
Gu i dance Journa I 1963, 4J_, 480-485.
American Psychological Association, Education and Training Board.
Criteria for evaluating training programs in clinical and counsel
ing psychology. American Psychologist, 1958, _P3, 59-60.
American Psychological Association, Subcommittee on Counselor Trainee
Selection, Counselor Training Committee, Division of Counseling
Psychology. An analysis of practices in counselor trainee selec
tion. Journal of Counseling Psychology, 1954, 1, 174-179.
Anderson, R. P., & Anderson, G. V. Development of an instrument for
measuring rapport. Personnel and Guidance Journal, 1962, 41, 18-
24.
Arbuckle, D. S. Client perception of counselor personality. JournaI
of Counseling Psychology, 1956, _3, 93-96.
Arbuckle, D. S., & Wicas, E. A. The development of an instrument for
the measurement of counseling perceptions. Journal of Counseling
Psychology, 1957, 4_, 304-310.
Armstrong, J. C. Perceived intimate friendship as a quasi-therapeutic
agent. Journal of Counseling Psychology, 1969, J_6, 137-141.
I 10


52
in counseling and students not in counseling with regard to quality of
their relationships with intimate friends and the tendency to use
intimate friends as therapeutic agents. In a very recent study Loesch
(1975) used the Relationship Inventory to assess whether a client's
nonverbaIized feelings and the counselors' perception of them during
the interview were related to subsequent counselor and client percep
tions of the counseling relationship.
The appropriateness of using the separate relationship dimensions,
purported to be measured by the Relationship Inventory, rather than the
more global total score, has been the subject of some controversy.
Initially Barrett-Lennard (1962), in asserting that separate relation
ship aspects were indeed being measured, stated that the scales were
not measuring a general factor, such as the client's overall satisfac
tion or dissatisfaction with the therapy relationship. However, Mills
and Zytowski (1967), on the basis of a princi pa I-components analysis
of the intercorrelations between the subtests for several samples using
a variety of forms of the Inventory, concluded differently. They found
that a general component accounted for about two-thirds of the variance.
WaIker and Little (I 969) chalenged the findings of Mills and
Zytowski since their conclusions were based upon an unrotated principal-
components analysis. Walker and Little then combined the Inventory
scores of 150 university undergraduates and conducted a component
analysis of the 64 items. They found three factors which they labeled
"non-evaIuative acceptance," "psychological insight," and "I ikabiIity."
At about the same time Tosi, Frumkin, and Wilson (1968) did inter-
correlational studies on the subtests of the Relationship Inventory
scores of 69 clients at the Kent State University Guidance Laboratory.


BIOGRAPHICAL SKETCH
Sidney Leon Bradley was born in Pembroke, Georgia, on September
22, 1942, the second child of Samuel and Ida Mae Bradley. He graduated
from Bryan County High School in May, I960. He received the Bachelor
of Arts degree in theology from Berkshire Christian College in 1964.
That same year he married Barbara Ann Sherrill, and he and his wife
moved to Friendship, Maine, where he became minister of the Advent
Christian Church.
In 1967 he enrolled in the University of Georgia School of Social
Work and received the Master of Social Work degree in 1969 with a
specialization in psychiatric settings. From 1969 to 1972 he was
Director of the Children's Program at the Advent Christian Home. In
1972 he began graduate work in Counselor Education at the University of
Florida as a Graduate Council Fellow. In October, 1973, he received
a faculty appointment in the Department of Psychiatry, University of
Florida, after completing his doctoral practica in the Adult Outpatient
Clinic. He received the Specialist in Education degree in Counselor
Education at the University of Florida in June, 1974. He simultaneously
completed his doctoral internship in the Adult Outpatient Clinic with a
subspecialization in marriage and family counseling while serving there
as Director of Group and Family Therapy.
In February, 1976, Mr. Bradley returned to the Advent Christian
Home in Dowling Park, Florida, with his wife and two sons, Jonathan and
128


THE RELATIONSHIP BETWEEN RESPONSE PREFERENCES
AND COUNSELOR EFFECTIVENESS RATINGS
OF COUNSELOR TRAINEES
By
SIDNEY LEON BRADLEY
A DISSERTATION PRESENTED TO THE GRADUATE COUNC
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR
DEGREE OF DOCTOR OF PH ILOSOPHY
L OF
THE
UNIVERSITY OF FLORIDA
1976

Ded¡cated to
Barbara Sherrill Bradley

ACKNOWLEDGMENTS
The author wishes to express sincere appreciation to the following
people who assisted in this study:
Dr. Joseph Wittmer, the chairman of the committee, whose leader
ship, experience, and precious time were invaluable assets.
Dr. Larry C. Loesch whose expertise in research design and statis
tical analysis was essential to this study and deeply appreciated.
Dr. Paul G. Schauble whose sensitivity, encouragement, and insight
were major contributions.
Dr. James L. Lister whose instrument provided the primary focus of
this study and who originally served as the writer's committee chairman.
Ms Barbara Rucker whose skills contributed both to the computer
analysis and to the splendid job of typing this study.
Barbara, my wife, whose love, patience, and understanding were
primary sources of inner strength during my entire graduate school
experience. It is to her that this study is appreciatively and affec
tionately dedicated.
ui

TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS ¡
TABLE OF CONTENTS iv
LIST OF TABLES vi
ABSTRACT v i i
CHAPTER I INTRODUCTION I
Purpose of the Study I
Rationale for the Study I
Statement of the Problem 4
Hypotheses 5
Definition of Terms 5
CHAPTER II REVIEW OF RELATED RESEARCH 7
Approaches to Counselor Effectiveness 7
The Speculative-Philosophical Model 8
The Psychometric Model 10
The Criteria Crisis in Effectiveness Model Building ... 13
The Communications Analysis Model 17
The Perceptual Model 19
The Carkhuff-Truax Therapeutic Conditions Model 23
Summary 27
Counselor Effectiveness Ratings 28
Supervisor Ratings 30
Client Ratings 31
Summary 34
CHAPTER I I I METHODOLOGY 36
Description of the Sample 36
Data Col lection 38
Limitations 41
Instrumentation 42
Helper Response Preference Inventory (HRPI) 42
Barrett-Lennard Relationship Inventory (BLRI) 46
Counselor Evaluation Rating Scale (CERS) 53
Analysis of Data 57

CHAPTER IV RESULTS OF THE STUDY
60
Introduction 60
Descriptive Data of the Sample 60
The Counselor Trainees 60
Supervisors 63
Cl ients 63
Client Evaluation Scores on the Counselor Trainees 63
Supervisor Evaluation Scores on the Counselor Trainees .... 64
Scores of the Counselor Trainees on the Predictor
Instrument 65
Relationships among the Three Instruments 66
Testing the Hypotheses 72
Test of the First Major Hypothesis 72
Test of the Second Major Hypothesis 74
CHAPTER V DISCUSSION AND SUMMARY
Summary of the Results 80
Practical Limitations 81
Interpretation and Discussion of Findings 83
Appl ¡cation 85
Suggestions for Further Research 87
Conclusions of the Investigation 88
APPENDICES 90
A Counselor Trainee Demographic Data Sheet 91
B Cover Letter to the Client 92
C Relationship Inventory Form OS-M-64 93
D Relationship Inventory Form 0S-F-64 96
E Counselor Evaluation Rating Scale 99
F Helper Response Preference Inventory 101
BIBLIOGRAPHY NO
BIOGRAPHICAL SKETCH 128
v

LIST OF TABLES
Table Page
1 Split-Half Reliabilities for Three Samples: Graduate Level
Counselor Trainees, Paraprofess ionaI Counselor Trainees
in a Community College, and ParaprofessionaI Disaster
Relief Counselor Trainees 44
2 Factor Labels and Factor Scores for the Helper Response
Preference Inventory 46
3 Test-Retest Reliability of the Subtests of the 64-1 tern
Relationship Inventory in Several Studies 49
4 Demographic Data on the Participating Counselor Trainee
Subjects 62
5 Means, Standard Deviations, and Ranges of Scores on the
Barrett-Lennard Relationship Inventory (N=46) 64
6 Means, Standard Deviations, and Ranges of Scores on the
Counselor Evaluation Rating Scale (N=46) 65
7 Means, Standard Deviations, and Ranges of Scores on the
Helper Response Preference Inventory (N=46) 66
8 I ntercorrelat ion Matrix of All Variables of the Barrett-
Lennard Relationship Inventory, the Counselor Evaluation
Rating Scale, and the Helper Response Preference
Inventory 67
9 Mean Correlation of Each Helper Response Preference
Inventory Factor with All of the Scores of the Two
Criterion Instruments 71
10 Barrett-Lennard Relationship Inventory Predictors 73
11 Counselor Evaluation Rating Scale Predictors 75
12 Regression Equation for the Barrett-Lennard Relationship
Inventory Predictors 77
13 Regression Equation for the Counselor Evaluation Rating
Scale Predictors 78

Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy
THE RELATIONSHIP BETWEEN RESPONSE PREFERENCES
AND COUNSELOR EFFECTIVENESS RATINGS
OF COUNSELOR TRAINEES
By
Sidney Leon Bradley
December, 1976
Chairman: Joseph Wittmer
Major Department: Counselor Education
This study investigated the relationship between response preferences
of counselor trainees and effectiveness ratings by their supervisors and
clients. A total of forty-six counselor trainees, enrolled in a graduate-
level practicum or internship, participated as subjects in this research.
Each counselor trainee completed the predictor instrument, the Helper
Response Preference Inventory (HRPI), which yields ten factor-analyzed
sub-scales. Client perceptions of the relationship were obtained by
having the trainee rated by a chosen client on the Barrett-Lennard Rela
tionship Inventory (BLRI). The supervisor criterion instrument was the
Counselor Evaluation Rating Scale (CERS) on which each trainee was
evaluated by an individual supervisor.
It was hypothesized that a relationship would be found between
response preferences on the HRPI and client evaluations on the BLRI.
It was also hypothesized that a relationship would be found between HRPI
responses and supervisor evaluations on the CERS. The forward stepwise
vu

regression technique was used to analyze the data. Four demographic
variables: age, graduate grade point average (GPA), number of graduate
quarters completed, and number of quarters of supervised counseling
practica completed, were also entered into the regression equations.
The best predictor of the BLRI Total score was grade point average
followed by responses that were premature judgments. The BLRI Level of
Regard subscore was predicted best by frank feedback responses. The
second best predictor was GPA followed by empathic responses and then
responses from an external frame of reference. The BLRI Empathy Level
subscore was predicted best by GPA, but the scale was also predicted by
premature judgments. For the BLRI Unconditionality of Regard subscore
GPA was the strongest predictor, but frank feedback was a significant
negative predictor. Only GPA predicted the BLRI Congruence subscore.
Other than GPA the demographic variables were not significant predictors.
The only statistically significant predictor of the CERS Total score
was the set of responses which used clarification primarily. The
clarification response factor also was the best predictor of the
Counseling subscore of the CERS although the scale was also predicted
by responses that validate or affirm the counselee's identity. There
was no predictor of the CERS Supervision subscore. No demographic
variable predicted any of the CERS scores.
It was noted that no HRPI factor predicted successfully both BLRI
and CERS scores. Also the canonical correlation between any two of the
three instruments in the study was not found to be significant. Inter-
correlations between the instruments were also generally low. Never
theless, neither hypothesis, that a relationship exists between the HRPI
and each of the two criterion instruments, could be rejected except in
part.
viii

The investigator concluded that the HRPI predicts both client and
supervisor evaluation scores on the criterion instruments in differing
ways. Self-disclosing responses appeared to predict higher client
ratings while clarifying responses tended to predict higher supervisor
ratings in this study.
IX

CHAPTER I
INTRODUCTION
Purpose of the Study
The measurement of counseling effectiveness and prediction of
counseling'success are both central concerns to the counseling pro
fession in the era of accountability. Efficiency and professional
responsibility dictate the need for valid ways to evaluate competence
and to differentiate levels of effectiveness. This study has explored
the relationship between the effectiveness ratings of counselor trainees
and their characteristic ways of responding to clients as a possible
method of evaluating effectiveness. More generally, this study has
investigated the theory that a counselor's communication within the
helping relationship is related to his level of effectiveness as a
helping person. This study also attempted to specify measurable compo
nents of communication which distinguished more effective counselor
trainees in a graduate level training program from less effective
counselor trainees.
Rationale for the Study
Research that leads to identification of effective counseling is
clearly relevant to the process of counselor selection (ACES, 1967),
evaluation (Stablein, 1962), and prediction of future counseling behav
ior (Walton & Sweeney, 1969). Nevertheless, investigations which have
attempted to specify and predict effective counseling have frequently
I

2
been disappointing and inconclusive. One result has been that admission
criteria to counselor training has been severely criticized for showing
little relationship to effective counseling (Rapaport, 1968; Thoresen,
1969). These indictments appear warranted in light of the research which
suggests that counseling can indeed be "for better or for worse" (Car-
khuff, 1966; Lister, 1970). A mounting body of research indicates that
unless a counselor can embody personal fac Iitative qualities such as
empathy, warmth, genuineness, and concreteness, he may adversely affect
the client (Carkhuff & Berenson, 1967). The profession is challenged
by these findings to devise and employ more effective selection criteria
to increase the probability of a counselor making a beneficial rather
than a detrimental impact on the client (Thoresen, 1969).
Intellectual attributes have been used extensively to screen
candidates for counselor training programs. Unfortunately, while
ability measures may predict graduate grade point average with reason
able accuracy, neither aptitude test scores nor grade point averages
yield a significant correlation with counseling effectiveness (Car
khuff, Piaget, & Pierce, 1968; Bergin & Solomon, 1968). The Graduate
Record Examination, although widely used in graduate counselor educa
tion programs for selection purposes, appears to provide little meaning
ful predictive validity (Wittmer & Lister, 1971).
Since 1963, when the American Personnel and Guidance Association
encouraged a shift to the evaluation of personality variables in
counselor candidates, a vast array of personality correlates with
counseling effectiveness have been examined. Illustrative of this
attention to the personality of the counselor were the reviews of
Burnett (1954),Stoughton (1957), Hill and Green (I960), Stripling and

3
Lister (1963), Froehlich (1951), Cash and Munger (1966), and Whiteley
(1969). Again, the results have been conflictual and inconclusive.
In his review of research in this area, C. H. Patterson concluded that
the data were sporadic, often poorly integrated with the basic issue
questions, and generally inadequate (1967).
An entirely different area of investigation began in 1950 with the
publication of the book Therapeutic Counseling by E. H. Porter. His
book presented the thesis that the counselors response pattern tends
to enhance or impede the flow of helpful communication. He contended
that the responses a counselor selects from his broad range of options
have implications for his ability to be an effective helping person.
His assertion seemed to imply that the primary index of a counselor's
ability to interact helpfully was the skill with which he could communi
cate with another human being. This appears to hold today.
In a similar vein, an entire school of counseling and psycho
therapy has developed around communication as the central concept in
human interaction (Bateson et al., 1956; Jackson, 1965; Haley, 1963;
Satir, 1967). While communication involves much more than simple
verbal exchanges, the communication theorists emphasize the importance
of the "report" or "message" component. Their belief is that largely
through verbal responses to himself, or "reflected appraisals" as
Sullivan (1953) called them, a person constructs his self concept, his
conception of the outside world including percepts of other persons,
and his sense of identity based upon how he personally fits into that
outside world. These theoretical constructs have been supported in
part by the process research on counseling which has linked at least
four variables to beneficial client outcome (Carkhuff & Berenson,

4
1967). These variables, empathy, warmth, genuineness, and concrete
specificity appear to be integrally related to, and generally are
measured by, the degree of sensitive, appropriate responsiveness
demonstrated by the person in the counselor role.
The question which naturally emerged from the above was whether
pure communicationaI response style was sufficient to identify and pre
dict effective counseling. For example, if Porter's assumption that
understanding, interpretive, etc. responses facilitate beneficial
client change were empirically valid, then perhaps an investigation of
response preferences could reveal one crucial factor in effective coun
seling. In fact, it might be possible to provide stimuli similar to
the counseling situation and test on paper how a counselor candidate
or trainee would prefer to respond in some future counseling relation
ship. This study explored whether those assumptions were warranted in
reference to one particular instrument for measuring response preferences.
Statement of the Problem
This study has investigated whether or not a measurable relation
ship exists between the effectiveness ratings of a counselor in training
and certain interpersonaI communicationaI modes. The question was
examined utilizing two discrete criteria. One criterion related
preferred verbal responses to client perceptions of the relationship
shared with the counselor trainee. The second criterion explored the
relationship of counselor trainee response preferences to the evaluation
of effectiveness made by the counselor trainee's individual supervisor.
tore specifically, the questions answered by this study are:
I. Does a counselor trainee's response preferences on a printed

5
counseling-simulation instrument have a significant relation
ship to clients' perceptions of a beneficial relationship?
2. Does a counselor trainee's response preferences on a printed
counsel ing-simuI at ion instrument have a significant relation
ship to supervisors' evaluations of counseling effectiveness?
Hypotheses
The general hypothesis in this research was that a relationship
would be found between the response preferences of counselor trainees
on the experimental instrument and two separate measures of counseling
effectiveness; namely, supervisor ratings and client perceptions of the
relationship. The following sub-hypotheses were tested:
HO| There is a relationship between counselor trainee response
preferences and client evaluations of their counseling
effectiveness.
H02 There is a relationship between counselor trainee response
preferences and supervisor evaluations of their counseling
effectiveness.
Definition of Terms
In this study the following definitions of relevant terms apply:
1. Counselor Response Preference The choice of a verbal reply to a
client's stimulus statement. In this study the helpee statement on
each item of the Helper Response Preference Inventory was the
stimulus. The counselor trainee was asked to choose one of two
helper responses for his reply. Since the format was forced-choice
the range of options was limited to only two response possibilities
2. Counselor Trainee Graduate students in Counselor Education

6
pursuing either a specialist or doctoral degree at the University
of Florida. Each trainee was currently engaged in either a super
vised practicum or internship in counseling at the time of data
coI Iection.
3. Client Ratinqs The perception by a client of the helpful quality
of the relationship with the counselor trainee. Operationally, the
client rated the counselor trainee on the Barret-Lennard Relation
ship Inventory. Higher scores were assumed to be indicating of a
more helpful relationship in the perception of the client.
4. Supervisor Evaluations The perception of the counselor trainee's
effectiveness by his individual supervisor. The ratings were
operationalized as scores on the Counselor Evaluation Rating Scale.
Higher scores indicated the supervisor perceived the counselor
trainee to be relatively more effective as a helping person.
5. Counseling Effectiveness The abiIity to a counselor trainee to
beneficially affect a client. For the purposes of this study,
counseling effectiveness was operationally defined in terms of
ratings of the counselor-in-training by clients and supervisors.

CHAPTER I I
REVIEW OF RELATED RESEARCH
The review of literature focuses on the following areas: I) the
varied approaches to conceptualization and measurement of counselor
effectiveness, and 2) the application of supervisors' and clients'
ratings to counselor effectiveness research.
Approaches to Counselor Effectiveness
The pursuit of a consistently viable way to assess the effective
counselor has often led to confusing and contradictory results. Quite
early, the sentiment had crystallized, and was later traced historically
by Polmantier (1966), that the profession needed to recruit counselor
candidates with personal qualities which would enhance their success as
counselors. Subsequently, the process of identification of those
essential personal qualities has met with mixed results. In this review,
five primary approaches to the search for a definition of counselor
effectiveness are explored: The speculative-philosophical approach, the
psychometric approach, the communication analysis approach, the percep
tual approach, and the core dimensional approach derived from process
research. After the first two approaches are presented, the question of
the criterion problem in effectiveness research is discussed to intro
duce the consideration of the final three approaches to counselor
effectiveness research.
7

8
The Speculative-Philosophical Model
The most primitive approach to establishing the components of
effective counseling began with the emergence of serious attempts to
conceptualize the necessary ingredients. In the widely quoted Fiedler
(1950) study it was concluded that experienced counselors, regardless
of frame of reference, agreed on the fundamental characteristies of an
ideal counselor-client relationship. The counselors concurred that the
ideal therapeutic relationship involves I) an empathic relationship,
2) rapport between client and counselor, 3) the counselor staying with
the client's problems, 4) the freedom to self-disclose, 5) mutual trust
and confidence, 6) excellent rapport, 7) active client participation,
8) freedom to make choices by client, 9) acceptance of client feelings
as normal and understandable, 10) a tolerant atmosphere, and II) a
feeling of being understood on the client's part.
Concurrent with the analysis of the ideal relationship was a
growing interest in those personal qualities of the counselor which
engender and enhance this therapeutic relationship. In their review of
this early literature, Barry and Wolf (1958) concluded that the
philosophical approach in this era consisted primarily of cataloguing
assorted virtues to be sought in prospective counselors. Graves (1944)
specified integrity, vitality, judgment, industriousness, high personal
standards, adaptability, experience and training. Cox (1945) added such
qualities as fairness, sincerity, good character and wholesome philosophy.
In addition to the above, Bowler and Dawson (1948) emphasized maturity,
respect, resourcefulness, reliability, a sense of humor, and an ability
to listen and keep confidences. Then in 1949 the National Vocational
Guidance Association proclaimed that counselors, ideally, were interested

in people, patient, sensitive to others, emotionally stable, objective,
respectful of facts, and trusted by others.
9
Hamrin and Paulsen in 1950 attempted to introduce a slightly more
scientific approach. They conducted a field study of 91 counselors to
arrive at a consenses of desirable traits which they believed facilitated
counseling. In descending order of frequency they listed I) under
standing, 2) sympathetic attitude, 3) friendliness, 4) sense of humor,
5) stability, 6) patience, 7) objectivity, 8) sincerity, 9) tact, 10)
fairness, II) tolerance, 12) neatness, 13) calmness, 14) broadminded
ness, 15) kindliness, 16) pleasantness, 17) social intelligence, and
18) poise.
By I960, however, Hill and Green reported that the enthusiastic
search for the essential personality pattern of a counselor had all but
dwindled away. The focus by then had shifted from speculative lists of
character traits to examining the total personality of the counselor
and how that personality impacts upon the counseling relationship. This
transition naturally arose out of a growing disenchantment with the
subjectivity of the philosophicaI-specuI at ive approach (Cottle, 1953;
Hill and Green, I960; Jones, 1951). While no massive revolution in the
understanding of effective counseling resulted from the philosophical
pursuit of ideal counselor characteristics, such outgrowths as Rogers'
formulation (1957) concerning empathy, genuiness, and unconditional
regard proved to be of considerable heuristic value. Cottle (1953)
recognized the helpfulness of the trait listings but criticized them
because the lists (I) were merely opinions, (2) failed to distinguish
counselors from related professionals, (3) varied widely, and (4)
failed to specify which patterns or interreI at ions of traits were

10
important. Consequently, the pursuit ot measurable personality
variables of effective counselors began in earnest using more objective
measures.
The Psychometric Model
The use of psychometric test data to link personal variables with
counseling effectiveness has been a popular, broad-based movement.
Actually a number of researchers have been able to effectively discrimi
nate quality levels in counseling using personality measures. However,
in general, the differentiations between more and less effective
counselors have not succeeded in identifying a dependable instrument
for screening. The principal reasons why such an instrument has not
been found are that positive findings are seldom replicated and results
add up to be in consistent and inconclusive. In general, the psycho
metric approach has encompassed personality, interest, and attitude
measures of counselor effectiveness. To simplify the pro Iiferation of
research in this area, this review will examine several popular measures,
one at a time.
The Edwards Personal Preference Schedule (EPPS) has been widely
used as a personality instrument to discriminate effective from less
effective counselors. However, the results overall have been per
plexing. Steffi re, King, and Leafgren (1962) used the Edwards Personal
Preference Schedule to examine participants in a one semester institute.
They found "most chosen" counselors, on the basis of potential, scored
higher on "deference" and "order" and lower on "aggression" and "abase
ment" than the "least chosen" counselors. On the other hand, the
counselors judged to be more effective in the study by Truax, Silber,

and Wargo (1966) were found to have higher scores on "change" and
"autonomy" and lower scores on "order." Oh I sen (1967), also using the
Edwards Personal Preference Schedule, measured two separate sample groups
of counselors. In the first group the most effective counselors scored
higher than the least effective on "succorance" and lower on "order,"
"deference," and "consistency." In contrast, Ohlsens most effective
counselors in the second group scored higher than the least effective
on "intraception" and lower on "dominance" and "aggression." Demos and
Zuwaylif (1966) found the EPPS to be the only one of three psychometric
measures to discriminate accurately among 30 secondary school counselors
in an NDEA institute based upon "multiple objective and subjective
criteria." Their most successful counselors scored significantly higher
on "nurturance" and "affiliation" while the least successful group
scored significantly higher on "autonomy," "abasement," and "aggression."
It is apparent that despite the similarities overall, the general
pattern of results is much to inconsistent to permit the broad scale
use of the EPPS as a reliable predictor of effective counseling.
Another standardized personality instrument which has been applied
in attempts to measure counselor effectiveness is the 16 Personality
Factor Profile (16 PF). In 1968 a regression equation was derived by
McClain for predicting counselor effectiveness from 16 PF scores. The
limited use of this prediction equation may be related to the generally
significant but inconsistent results found in early research attempts
to use the 16 PF to differentiate levels of effectiveness (Donnan,
Harlan and Thompson, 1969; McClain, 1968; Myrick, Kelly, & Wittmer,
1971; and Wittmer & Lister, 1971).
The Rokeach Dogmatism Scale has been used with considerable promise

12
in this area. Walton and Sweeney (1969) state that in their review
they found that "results of work utilizing this instrument have rather
consistently supported its capability for distinguishing between coun
selors on opposite ends of the effectiveness continuum" (p. 34).
Steffi re, King, and Leafgren (1962) found, among counselor trainees
selected as having more counseling potential by their peers in an MDEA
Guidance and Counseling Institute, significantly less dogmatism than,
among those trainees not chosen. Cahoon, in 1962, also reported that he
had found significantly less dogmatism in superior counselors. These
findings have been supported by Kemp's finding (1962) that low scores in
dogmatism on the Dogmatism Scale were related to openness in an actual
counseling situation. Still further corroboration came from the conclu
sion by Russo, Kelz, and Hudson (1964) that openmindedness is indeed an
important quality in a counselor after they found 12 items on the
Dogmatism Scale which distinguished significantly between counselors
judged most and least effective. Furthermore, in studies that did not
use the Dogmatism Scale but measured a closely related factor, tolerance
for ambiguity, both Brams (1961) and McDaniel (1967) found a relation
ship between this factor and high ratings of effectiveness.
Virtually a myriad of other personality measures have been tried in
attempts to isolate personality patterns which appear to coincide with
effective counseling. Some illustrative efforts have included score
profiles from the Miller Analogies Test and the MMPI (Abeles, 1958),
the MMPI alone (Brams, 1961; Foley & Proff, 1965), the Strong Vocational
Interest Blank (Stefflre, King, & Leafgren, 1962), the Kuder Preference
Record (Blocher, 1963), the Wisconsin Relationship Orientation Scale
(Wasson, 1965), the Al Iport-Vernon-Lindzey Scale of Values (Demos &

13
Zuwaylif, 1966), the Self-Description Checklist (Wicas & Mahan, 1966),
the California Personality Inventory and the MMPI (Schroeder & Dowse,
1968), and the Interpersonal Check List (Eberlein & Park, 1971). While
some of these studies have reported promising avenues for reliable
prediction, frequently the attempts to replicate have either yielded
conflictual findings or qualified the original results in ways that
diminish confidence in the ultimate dependability of the instrument.
An excellent illustration of this frustrating process has been the
attempts to differentiate the measurable characteristics of A and B
type therapists as first described by Whitehorn and Betz (1954, I960).
They found that type A therapists who established warm, personal rela
tionships had substantially better success rates with schizophrenics
than the more passively detached and interpretive type B therapists.
Furthermore, they found that 23 items on the Strong Vocational Interest
Blank distinguished the two therapist types. However, McNair, Callahan,
and Lorr (1962) found that with neurotic patients, and psychologists as
therapists, the results were reversed and the type B therapists were
more successful. Carson et al. (1964) concluded from their research
that the type A therapist works best with distrusting clients and type
B with trusting clients, in an attempt to reconcile the discrepancies.
Kemp (1966) introduced still more controversy by concluding from his
research that both types of therapists are most capable of helping those
clients who produce the most discomfort in themselves. It is readily
apparent that the findings are intriguing but decidedly inconclusive.
The Criteria Crisis in Effectiveness Model Building
The disappointing attempts to link measurable personality variables

14
to counseling effectiveness is mystifying. It is enigmatic why results
overall have been either contradictory or confusing. One criticism of
this entire approach came very early from Stern, Stein, and Bloom (1956)
They felt the inconsistency eventuated from the inadequacy of the trait
factor philosophy in which the approach was rooted. They called for
attempts to identify meaningful personality traits related to counseling
behavior only after predictions have been derived from we I I-art¡cu Iated
theories.
Other observers take a different perspective and view the problem
as an artifact of the "effectiveness" criteria invoked. The profession
of counseling has not attained a consensus on the goals of counseling,
the differential counseling methods appropriate to specific problems,
or standard indicators of client progress. Therefore, the elusive con
cept of "counselor effectiveness" may well be relative not only to the
uniqueness of the counselor, but also to the setting, the problem, and
the client (Stefflre, King, & Leafgren, 1962).
Steffi re, King, and Leafgren (1962) pointed out that we are unable,
at our present knowledge level, to reach consensus about how to identify
or even define the effective counselor. Nevertheless, societal pressure
urges the profession to educate "as if" the "effective counselor" was a
definable entity. Generally, ratings of supervisors or judgment by pro
fessional practitioners has been used to differentiate levels of skill,
but client ratings have gained in popularity and are explored in depth
in the next section of this chapter. However, the ratings utilized have
proliferated in effectiveness studies to include peer judgments, self-
ratings, etc., which has added to the profusion of confusion.
Researchers have attempted at times to look at behavioral criteria

15
or "unobtrusive measures" with a variety of behaviors being studied.
Evaluation of counseling success has varied from dropout rate to atten
dance and grades to the proportion of time spent talking by the client.
A serious question, however, has been raised by Hill (I960) regarding
the validity of such outcome criteria. In reality such behavioral
criteria are responsive to change effects from a host of uncontrollable
and often unidentifiable variables.
In their I960 review of research, Hill and Green traced a large
portion of the difficulty in finding a consensual definition of effec
tiveness to the complexities of the settings within which counselors
function. They cited the Iiterature to support their contention that
the issues of selection and effectiveness are often confused. Large
variations in roles and relationships exist even in one counselors work
(APA, 1958; Hoffman, 1959), and across jobs the diversity of skills and
attitudes needed by counselors becomes staggering. There is growing
awareness that often, as Hill and Green (I960) assert, the counselor may
be called upon to fulfill responsibilities which the profession has
neither embraced nor prepared the counselor to do (Purcell, 1957;
Wellman, 1957). Today the role and function of the counselor appears to
be diverging even more from a precisely defined model which is amenable
to unambiguous evaluation.
Beyond the lack of agreement as to what is the preferred model of
effective treatment for a specific clientele, lies the diversity of
opinion about the different schools of therapy. In one study Sunderland
and Barker (1962) found that experienced therapists from Freudian,
Sullivanism, and Rogerian schools exhibited different behaviors, but
precisely how the specific counseling approach affects the effectiveness

16
of the counselor in the wide range of possible settings,
'~5, and
problems is obscure.
Still we have not exhausted the variables that may cc'fc
,'d an
attempt to define and measure an effective counseling rela~'c
-shi p.
Some variables that have gained attention are: the attit-Si'
a 1 and
value system of the therapist (Combs & Soper, 1963; White's"
5 Betz,
1954); demographic variables (Bailey, Warshaw, & Eichler, 195
9); the
experience of the therapist (Cartwright & Vogel, I960); the c
scree of
disturbance of the patient (Stone, Frank, Nash, & Imber, 196
; and the
clients felt need to change (Cartwright & Lerner, 1963). Ob
.iously
this list is not comprehensive, but it clearly illustrates rr
s host
of variables which have been found to influence therapeutic e
--ective-
ness. If all the variables which it is possible to study .-.er
s added to
the list of those which have been examined, the resulting cc^
scsite
would likely be astronomical.
For some the complexity of teasing effectiveness out c~
"is laby-
rinth of variables is felt to be overwhelming. A few cone _s
s _nat
counseling effectiveness research is so elusive as to ever ce
ssyond
critical inquiry (Thoresen, 1969). Some writers feel the
:-erica1
foundations of counseling have been shaken (Ford & Urban, ¡96
, and
others fear the super-structure of the profession is dangers-
s / close
to collapse into chaotic confusion and revolution (Colby, 95^
;; boltz, 1966; Rogers, 1963). Simultaneous with this expresses
concern
is the growing pressure for the profession to be increasi nc :
account
able. The ground swell of public demand calls for the course
's pro-
fession to demonstrate selection, evaluation, and education c-
'icedures
which are derived from and justified by a substantial body e~
e-~s i ri ca 1
research on effectiveness (Thoresen, 1969).

17
Perhaps the most viable alternative is for the profession to move
more toward the "systems approach" to conceptualizing and evaluating
counseling. A key concept is that the system is more than the sum of
its parts. For example, the counseling encounter Is a total relation
ship built upon two distinct but temporally engaged personalities
attempting to bridge their separateness via a communications process.
The crucial variables may be the unique perceptual worlds of the coun
selor and client, the communication that flows between them, and the
emotional tone of the atmosphere within which they transact. Perhaps
this triad offers a viable and comprehensive approach to the enormously
complex give-and-take of the counseling process. Since the variables
under investigation in this study relate specifically to communication,
it is important to consider the relevance of this factor to counseling.
The Communications Analysis Model
Fiedler (1950, 1951) pointed out the crucial significance of the
relationship to counseling. He reported that experts excel non-experts
in their ability to communicate and understand their patients. Porter
(1950) went a step further and hypothesized that the counselor's style
of preferred responses sets the tone of the interview and defines the
counselor's capacity to be helpful. In Porter's instruments, the Porter
Test of Counselor Attitudes and the Porter Interview Analysis Scale, he
categorized responses as understanding, supportive, i nterpretive,
probing and evaluative. A number of studies of trainees in counselor
education programs have revealed that with exposure to training and
supervision trainees become more understanding and less evaluative,
supportive, and probing (Demos & Zuwaylif, 1963; Kassera & Sease, 1970;

18
Munger & Johnson, I960). A paradoxical finding, however, has been dis
covered among rehabilitation counseling trainees where preferences
are decidedly in favor of responses that inform, question, probe and
diagnose (Davis & O'Conner, 1974; Wittmer & Lister, 1971).
In his doctoral dissertation Chap line (1964) related level of cog
nitive complexity in counselors to preferences for understanding or
evaluative responses. He administered the Construct Repertory Test to
his sample of 163 graduate students in counseling and guidance from the
departments of psychology and education to measure cognitive complexity.
In addition to gathering demographic data, he had the subjects respond
in writing to an experimental tape recording of client statements.
Using Porter's original categories (1950), he found that subjects who
measured high on cognitive complexity were also high on evaluative
responses while the low cognitive complexity scorers were high on
understanding responses. The education students tended to prefer
evaluative responses and the psychology students leaned toward under
standing responses at a higher-than-chance level. On the other hand,
response preferences were unrelated to sex, age, level of training,
previous personal counseling, teaching experience, or even experience in
counseling. Freedman, Antenem, and Lister (1967) in a related study
found a strong predictable relationship between verbal response patterns,
as measured by Porter's Scales, and counselor personality characteris
tics on the California Psychological Inventory. The response-analysis
of their 15-minute roleplayed interviews was suggested as a useful
procedure for student selection and progress evaluation.
Grigg and Goodstein (1959) found that clients who saw their coun
selors as verbally active were much more likely to report a favorable

19
outcome than those who saw their counselors as passive listeners. In
his 1961 report Brams emphasized the strong tie between the counselor's
personality and his ability to communicate with clients, and Pal lone
and Grande (1965) even concluded that the counselor's verbal mode sig
nificantly influences the amount of the client's probIem-reIevant
communication. Clearly the counselor's communication style is funda
mental to the relationship established with the client. However, the
extent of this influence has not been established. This study, in part,
will be an attempt to evaluate the impact of verbal response preference
to a multi-factor measure of counseling effectiveness.
In the belief that the counselor's communicational mode is a direct
expression of one's unique perceptual world, and that in turn one's
communicational style crucially shapes the atmosphere of the counseling
relationship; the focus of this review now turns to the last two
approaches to counseling effectiveness relevant to this study: the
perceptual approach and the core therapeutic conditions approach as
epitomized in the writings of Carkhuff.
The Perceptual Model
A much different approach to the question of counselor effective
ness has been introduced by perceptual psychology. According to this
approach (Combs & Snygg, 1959), the internal perceptions and perceptual
organization of the individual provide a meaningful perspective for the
examination of effectiveness in human helping. The assumption is made
that counselors who have "learned to use themselves as effective instru
ments in the production of helping relationships can be distinguished
from those who are ineffective on the basis of their characteristic

20
perceptual organizations"(Combs & Snygg, 1959, p. 14), The perceptual
approach then may be contrasted with more behavioristic approaches which
operationalize the criteria of effective helping in terms of specifiable
behaviors.
According to the perceptual approach, a man's behavior is a func
tion of his "perceptual field," By perceptual field is understood the
individual's unique perceptions of himself and the world in which he
lives. Hence, the field consists of what the person perceives at a
given point in time and the meaning or significance he attaches to that
perceptual experience (Combs, Avila, & Purkey, 1971). Behavior is seen,
therefore, as a direct consequence of the individual's field of meanings
existing for him at the moment of behaving. Since the perceptual
psychologist is concerned with internal impression rather than external
expressions, the internal frame of reference is preeminent. Perhaps it
would not be an overstatement to say that the perceptual theorist
believes that a counselor is effective not because of his specific
behaviors but rather because of the internal perceptions behind those
behaviors.
Walton and Sweeney (1969) cite the study of counselor perceptions
by Combs and Soper (1963) as one of the more promising approaches to the
measurement of effectiveness. Using a critical incidents in human
relationships format, they found moderately high positive relationships
between 12 characteristic ways of perceiving and effectiveness as a
counselor. The effective counselor was sensitive to and concerned with
how things looked to others and was oriented to people rather than
things. He perceived others as able rather than unable, dependable
rather than undependable, friendly rather than unfriendly, and worthy

21
rather than unworthy. Moreover, he perceived himself as being identi
fied with people rather than apart from people, as personally adequate
rather than wanting, and as seif-revealing rather than self-concealing.
These dichotomous characteristics which Combs and Soper studied
grew out of a graduate seminar held in 1959 at the University of Florida
(Combs, 1961). The purpose of the seminar was to investigate effective
helping relationships from the internal point of view. The effort to
identify perceptual characteristics which were believed to be relevant
to the task of effective professional helping crystallized in five focal
areas (I) the general frame of reference from which the helper approaches
the situation, (2) the ways in which the helper perceived other people,
(3) the ways in which the helper perceived himself, (4) the ways in which
the helper perceived the task with which he was confronted, and (5)
the ways in which the helper perceived appropriate methods for carrying
out his purposes. The forty dichotomies issuing from that seminar are
listed below. Those perceptual dichotomies believed related to effec
tive helpers are on the left and those related to ineffective helpers on
the right.
(I) General Frame of Reference
InternaI
Growth orientation
Perceptual meanings
PeopIe
HopefuI
Causation oriented
Externa I
Fencing in or controlling
Facts, events
Things
Despairing
Mechanics oriented
(2) Perceptions of Other People Sees others as:
Capable
T rustworthy
HelpfuI
Unthreatening
Respectable
Worthy
Incapab I e
Untrustworthy
Hindering
Threatening
No account
Unworthy

(3) Perceptions of Self Sees self as:
Identified with people
Enough
T rustworthy
Li ked
Wanted
Accepted
Feels certain, sure
Feels aware
Self-reveaIing
(4) The Helping Task and Its Prob
Purpose is:
Helping
Larger meanings
Altruistic
Understanding
Accepting
Valuing Integrity
Approach to problem is:
Positive
Open to experience
Process oriented
Relaxed
Awareness of complexity
Tolerant of ambiguity
(3) Perceptions of Methods Sees
Apart from people
Wanting
Not trustworthy
Not Iiked
Not wanted
Not accepted
Doubt
Unaware
Se If-conceaIing
ems
Dominating
SmaI Ier meanings
Narcissistic
Condemning
Rejecting
Violating integrity
Negative
Closed to experience
Ends oriented
Compulsion to change others
Overs impI ification
Intolerant of ambiguity
appropriate methods as:
Helping
Cooperation
Asserting
Acceptance
Permissive
Open communication
Giving
Vital
Man ipuI ating
Competition
Appeasing
Rejecting (attacking)
Authoritarian
Closed communication
Withholding
Lifeless
(Combs, 1961, pp. 56-57)
More than twenty studies have attempted to infer the perceptual
organization of helpers using this variety of original perceptual dimen
sions plus a few that have been identified subsequently. Among the
more notable studies are those of priests (Benton, 1964), secondary
school teachers (Brown, 1970), counselor trainees (Combs & Soper, 1963)

23
junior college teachers (Dedrick, 1972), nurses (Dickman, 1967), college
professors (Doyle, 1969), elementary school teachers (Gooding, 1974),
residence assistants (Jennings, 1973), and counselors (Rotter, 1971).
The overall impression is that this substantial body of research suggests
a potentially rewarding means of identifying helpers with relatively
higher levels of helping ability.
The Carkhuff-Truax Therapeutic Conditions Model
The model of counselor effectiveness research which was popularized
by Truax and Carkhuff, and currently is often identified with the latter,
was formulated primarily out of the early research efforts of the pair
(Truax & Carkhuff, 1967). Their joint investigations were directed at
specifying the primary core of interpersonaI facilitative factors which
relate to favorable outcome in counseling and therapy. After extensive
research, the two concluded that the core dimensions of accurate empathy,
nonpossessive warmth, genuineness, and concreteness account for between
one-third to one-half the variability in client outcome (Truax, 1961;
Truax & Carkhuff, 1966). Counselor trainees who offer their clients
high levels of these core dimensions have been found to consistently
elicit beneficial therapeutic gain (Berenson, Carkhuff, & Myrus, 1966;
Carkhuff & Truax, 1965; Truax & Mitchell, 1971).
The work of Truax and Carkhuff is clearly an outgrowth of the
Rogers' formulations. Truax, a student of Rogers, developed a "Tenta
tive Scale for the Measurement of Depth of Self Exploration" along with
other scales rooted in the theoretical hypothesis of "necessary and
sufficient conditions" of therapeutic personality change proclaimed by
Rogers (1957). Much of the plethora of research summarized by Truax

24
and Carkhuff is based upon these scales. Carkhuff, after the two
researchers went divergent paths, modified each of Truax's therapeutic
conditions scales to incorporate five levels for each scale (Carkhuff,
1969a; Carkhuff & Berenson, 1967). The five levels each represent a
qualitative degree of therapeutic functioning, allowing quantification
of the extent to which a therapist corresponds with the standardized
criteria. At first the scales were used to rate excerpts of audio
taped counseling interviews. More recently video tapes have been the
medium under investigation (Kagan & Krathwohl, 1967), particularly in
the evaluation of the InterpersonaI Process Recall (I PR) technique.
Even though the reliability and validity of the Carkhuff scales are
at times questioned (Hefele & Hurst, 1972), the summarization of the
extensive body of research (Carkhuff & Berenson, 1967) presents one of
the most persuasively supported theoretical positions in counseling
today. Carkhuffs scales are probably the most widely used rating
scales in current studies of self-exploration and the facilitative con
ditions of empathy, warmth, genuineness, and concreteness.
A central concept in the model for effective counseling which is
presently being explicated by Carkhuff is that facilitative conditions
can not only be measured but can also be transmitted to counselor
trainees by counselor educators who themselves provide them (Carkhuff &
Truax, 1966). Since the accumulating weight of research supports the
belief that counseling can be "for better or for worse," depending
upon the presence or absence of these dimensions, the implications of
this research for counselor selection and accountability monitoring are
clearly evident. Since the factor subscales of the predictor instrument
in this research relate directly to the antecedent research of Carkhuff

25
and his associates, it seems appropriate to elucidate how he concep
tualizes the integration of the facilitative dimensions into the thera
peutic process. Two of his primary propositions and their corollaries
follow. These bold assertions are supported by research evidence in the
book Helping and Human Relations, Volume I (1969) from which they were
extracted.
Proposition I. The degree to which the helping person offers
high levels of facilitative conditions in response to the
expressions of the person seeking help is related directly
to the degree to which the person seeking help engages in
processes to constructive change or gain.
Corollary I. The degree to which the helping person offers
high levels of empathic understanding of the helpees
world is related directly to the degree to which the helpee
is able to understand himself and others.
Corollary II. The degree to which the helping person communi
cates high levels of respect and warmth for the helpee and
his world is related directly to the degree to which the
helpee is able to respect and direct warm feelings toward
himself and others.
Corollary III. The degree to which the helper is helpful in
guiding the exploration to specific feelings and content is
related directly to the degree to which the helpee is able
to make concrete his own problem areas.
Corollary IV. The degree to which the helper is responsively
genuine in his relationship with the helpee is related to the
degree to which the helpee is able to be responsively genuine
in his relationship with himself and others.
Proposition II. The degree to which the helping person
initiates action-oriented dimensions in a helping relation
ship is directly related to the degree to which the person
seeking help engages in processes that lead to constructive
change or growth.
Corollary I. The degree to which the helper can be freely,
spontaneously and deeply himself, including the disclosing
of significant information about himself when appropriate,
is directly related to the degree to which the helpee is
able to be genuine and self-disclosing in appropriate
relationships.
Corollary II. The degree to which the helper actively

26
confronts the he I pee and himself is directly related to the
degree to which the helpee is able to confront himself and
others.
Corollary III. The degree to which the helper both acts and
directs the actions of the helpee immediately in the present
to the relationship between helper and helpee is related to
the helpee's ability to act with immediacy and later to
direct the actions of others.
Corollary IV. The degree to which the helper can make con
crete a course of constructive action is related to the
degree to which the helpee can go on to make concrete courses
of action for himself and others, (pp. 84-90)
The complete text of Carkhuff's working assumptions emphasizes
the direct personal effect upon the helpee of the helper's level
of facilitative conditions offered in the interpersonal encounter.
It follows that the degree to which the counselor in the helping role
provides high levels of empathy, warmth, respect, genuineness, and
concreteness is directly related to the internalization of these
facilitative dimensions into his own life. Moreover, the degree to
which the helper is openly and deeply himself, self-disclosing, con-
frontive of himself and his client, present with immediacy, and active
with clear, concrete, action-oriented alternatives is again directly
related to the client's capacity to translate these same facilitative
dimensions into his own life sphere.
Carkhuff has written extensively about the selection process and
its relationship to the previously stated propositions. He has also
addressed himself to the effectiveness criteria problems already pre
sented. His position is that "the best index of a future criterion is
a previous index of that criterion" (1969b, p. 85). In general, the
counselor selection process must be distilled to identification of
specific functioning within the helper role by the prospective helper.

27
He suggests the following specific indices: I) an index of communica
tion; 2) an index of trainabiIity; 3) an index of self exploration; and
4) a critical incident index to ascertain his potential helpfulness in
the face of crisis. This study relates to the first and last of these
indices because the predictor instrument is a measure of communication
within a criticaI-incidents or counseling-simulation format.
Summary
Despite voluminous research with a myriad of approaches, the effec
tive counselor remains a nebulous concept. The speculative-philosophi
cal approach helped clarify for training purposes the ideal characteris
tics and proved thought-provoking for a period, but the approach came
under critical fire because its subjectivity did not solidify the base
of the profession.
Empiricism, in the form of psychometric testing, came into vogue
as attention focused upon the measurable personality traits of the
counselor. There was not found, however, many impressive links between
specific personality traits and beneficial client outcome. It appeared
that a more global assessment of the total impact of the helpers
personality rather than a trait-factor approach was more relevant to the
therapeutic involvement of a truly helpful relationship.
Perceptual dimensions have attracted considerable attention in the
counseling profession, but research on this approach has been less than
prolific. Focus upon the internal frame of reference of the counselor
has yielded some meaningful links with effective helping.
More recently, the pressures within and without the profession
toward accountability have led to attempts to define counseling

28
effectiveness in terms of demonstrable client outcome criteria. In
response, the Carkhuff model has endeavored to derive a coherent system
of selection and training based upon their conclusions from outcome
research. Nevertheless there remain the ubiquitous problems of effec
tiveness criteria, plausible indicators of therapeutic gain, consensual
counseling goals, and quantifying the variability in clients, settings,
and problems. While the ultimate goal has not been attained, the
impression is that greater definitional and procedural clarity have
evolved from the continual grappling with the problem of counselor
effectiveness criteria.
Counselor Effectiveness Ratings
The widespread use of global ratings of effectiveness as criteria
in counseling research has been previously mentioned. In their review
of the effectiveness literature, Walton and Sweeney (1969) point out
the broad base of acceptance rating scales have attained. Since the
original studies by Rogers and Dymond (1954) and the emergence of the
"Process Scale" conceptualized (1958) and published by Rogers (1959),
there have arisen a host of researchers and research instruments using
y I t. J y l 1 f J I S '. i -
a ratings format.
In addition to the Carkhuff scales, already cited, similar rating
scales include the Self-Anchoring Scale of Counseling Effectiveness
(Kilpatrick & Cantril, i960), the Communication Rating Scale (Brams,
1961), the Counselor Potential Scale (Dole, 1964), the Wisconsis Rela
tionship Orientation Scale (Wasson, 1965), the Counselor Evaluation
Inventory (Linden, Stone, & Schertzer, 1965), the Counselor Evaluation
Form (FauIkenberry, 1968), the Counselor Practicum Evaluation Form

29
(Eberlein & Park, 1971), and the Counselor Evaluation Rating Scale
(Myrick & Kelly, 1971), which was one of two criterion measures employed
in this research study.
There have been a variety of other unnamed rating scales including
a sorting process (Embree, 1954), and free response type counselor
perception instrument (Arbuckle & Wicas, 1957), sociometric indices
(Gade, 1967; Steffi re, King, & Leafgren, 1962), a categorized rating
instrument (Keltz, 1966), a sensitivity scale (O'Hern & Arbuckle, 1964),
and a number of semantic differential rating scales (Bartlett & Thomp
son, 1971; Redfering, 1973; Riese & Stoner, 1973; and Shapiro, 1968).
It can be said that rating scales have served a large number of
researchers as an accessible, simplified, and quantifiable approach to
the complex problem of finding a viable, valid effectiveness criterion.
In general, the above instruments were validated on the basis of accept
able correlations with other measures of counseling effectiveness such
as rankings or "expert" judgments. At times validity has been inferred
from reliability studies. While such practices are defended by respected
researchers in the field (Combs & Soper, 1963; Russo, Kelz, & Hudson,
1964; Steffi re, King, & Leafgren, 1962), it should be noted that actual
client outcome criteria were not generally used in the validation of the
above instrument.
Although the use of rating scales has extended into many areas that
are not germaine to this study, it can be noted that considerable
research has been generated using counselor self ratings, peer ratings,
coached client ratings, independent judge ratings, and a wide spectrum
of combinations. Nevertheless, this study focused specifically upon
the use of supervisor and client ratings as the basis for effectiveness

30
criteria. Both supervisor and client ratings are explored in some depth
in the review which follows.
Supervisor Ratings
There appears to be a general consensus that supervisor ratings are
among the most valid measures of effective counseling (Patterson, 1964).
The assumption is made that highly experienced counselors who comprise
"expert" panels are the most qualified of all judges to identify differ
ential levels of effective counseling. The logic of Cattell, penned in
1903, is still compelling today: "There is, however, no other criterion
for a mans work than the estimation in which it is held by those most
competent to judge" (p. 314). Clearly the determination of "those most
competent to judge" remains a debatable point, but Cattells rationale
continues to be invoked to justify the use of supervisor ratings.
Supervisor ratings have been frequently employed as a criterion
measure of counselor effectiveness. Such ratings were used in studies
by Allen (1967), Bishop (1971), Brown and Cannaday (1969), Demos and
Zuwaylif (1966), Embree (1954), Friesen and Dunning (1973), Gade (1967),
Keltz (1966), McDougalI and Reitan (1961), Watley (1967), Wicas and
Mahan (1966), and a host of others. For the most part interrater relia
bilities in these studies have been within the acceptable range. We
note, for example, that in the study by Gade (1967) the interstaff rank
ing achieved a reliability of .80 when computed by the coefficient of
concordance statistical formula. In their 1973 study Friesen and
Dunning found that supervisors had an interrater reliability of .94
on the Counselor Effectiveness Scale. This coefficient was higher than
the .92 achieved by lay people or the .91 reliability among practicum

31
students. Using the same instrument, Keltz (1966) found the interjudge
agreement on pooled ratings to range between .73 and .78 and concluded
that the instrument and the supervisor rating approach "constitute a
usable, realistic measure for assessing the counseling proficiency of
counselor-trainees" (p. 516). Likewise, when Falkenberry compared the
evaluation consistency of supervisor, peer, role-player, and self ratings
on the Counselor Evaluation Form, he concluded that for overall scores
and in specific items the supervisors were more reliable than any of the
other groups of raters.
A comprehensive review of published studies using the Counselor
Evaluation Rating Scale is presented in the next chapter. The CERS
(Myrick & Kelly, 1971) is one of the criterion measures employed in
this study and, therefore, is explored in depth under the section on
"Instrumentation."
Cl ient Ratings
The use of cIient perceptions to evaluate counseling effectiveness
has been extensive, and the reactions of researchers overall tend to
qualify the approach as a valid criterion. The value of client ratings
is attested to by a substantial number of authors (Arbuckle, 1956; Form,
1955; Grigg, 1961; Grigg & Goodstein, 1957, 1959; Linden, Stone, &
Shertzer, 1965; Mueller, Gatsch, & Ralston, 1963; Patterson, 1958;
Pohlman, 1961; Pohlman & Robinson, I960; Rosen, 1967; Severinsen, 1966;
Shoben, 1953; Stablein, 1962; Thompson & Miller, 1970).
An early cogent argument for the inclusion of client ratings in
effectiveness research was penned by Grigg and Goodstein (1957). Their
position was that:

32
Some appraisal of the client's reaction to the counselor and
to counseling should be obtained before we can say that we
have any comprehensive understanding of who makes a good
counselor ... a successful practitioner, among other things,
is one who elicits favorable reactions from the recipients
of his services. (Grigg & Goodstein, 1957, p. 31)
Indirect support for counselor ratings may be drawn from studies
which demonstrate significant positive correlations between client and
supervisor rankings of counselor effectiveness (Brown & Cannaday, 1969).
In several studies using the Counselor Evaluation Inventory (CEI) devised
by Linden, Stone and Shertzer (1965) high correlations between client
and supervisor ratings lend support to the hypothesis that clients and
supervisors tend to evaluate a similar set of criteria (Anderson &
Anderson, 1962; CorrelI, 1955; Poole, 1957). In research involving the
same instrument, Rickabaugh, Heaps, and Finley (1972) discovered a
significant relationship between clients' perceptions of counselor com
fort and client academic grade improvement, another frequently used
measure of effective counseling outcome. Pfeifle (1971) also found that
clients rated counselors with practicum experience significantly higher
than counselors without practical experience when ranked on the CEI.
Client ratings have often been an integral component in the pre
viously discussed research on the therapeutic conditions of empathy,
warmth, genuineness, respect, and concreteness offered by the counselor
(Truax & Carkhuff, 1967). One of the pioneers in using client percep
tions to measure the therapeutic conditions was Barrett-Lennard (1962).
An extensive analysis of the characteristics and uses of his instrument
follows in the section of the next chapter on "Instrumentation." The
use of his instrument and of similar relationship inventories has yielded
positive results with client samples ranging from juvenile delinquents

33
(Truax, Wargo, Carkhuff, TurnelI, & Glenn, 1966) to schizophrenics
(Rogers et al., 1967) to undergraduates seen at a university counseling
center (Kurtz & Grummon, 1972). The review of the broad scale applica
tions of the Barrett-Lennard Relationship Inventory should be noted in
Chapter III.
The issue has sometimes been raised that subjectivity and bias tend
to invalidate client perceptions as a valid criterion of counseling
success (Patterson, 1958, 1959; Pohlman, 1961; Rosen, 1967). One
criticism has been that unrealistic expectations may lead to inevitable
dissatisfaction with a consequent blurring of objectivity (Patterson,
1958; Severinson, 1966). Some support for this position came from a
study by Pholman and Robinson (I960) who felt their research confirmed
the existence of definite counselor preferences among clients. The study
consisted of verbal presentation of 92 counselor behaviors to a group of
psychology students who rated the behaviors in terms of degree of
annoyance or liking. Mean ratings of 73 behaviors differed signifi
cantly from the expected mean of neutrality. Furthermore, in a review
of the literature on client preferences, Rosen (1967) listed a substan
tial number of other client preferences which may influence a clients
perception and subsequent rating of counselor effectiveness.
Another variable, client preferences for specific counseling
approaches or sets of counseling techniques, may also alter the clients
ratings. Greater satisfaction was reported by Barahal, Brammer, and
Shostrom (1950) for clients counseled by "client-centered counseling"
than for clients counseled by "traditional" techniques. One must be
cautious, however, in interpreting such findings because a client's
phenomenon influenced by a variety of confounding and interacting

34
variables. Among these variables are the clients actual problem
(Bordin, 1955), the personality of the client (Sonne & Goldman, 1957),
the degree of counselor activity (Forgy & Black, 1954), the identity of
the counselor (Grigg & Goodstein, 1957), and perhaps a gamut of other
possible variables. Patterson (1959) pointed out that one pernicious
danger in interpreting client ratings is the possible discrepancy be
tween what a client wants and feels he needs and what is in the clients
best interest. It is conceivable that a client may rate a counselor
positively when in reality the counselor may be compounding the clients
problem. In fact, some studies have suggested that clients may not be
the most objective judges of the counselor's therapeutic conditions
(Burstein & Carkhuff, 1968; Truax, 1961).
In conclusion, the preponderance of research evidence supports the
use of client perceptions and ratings of counselor effectiveness.
Nevertheless, one must be mindful of the complex variables such as
preferences, expectations, and situational factors which may impinge
upon and qualify their meaning. Therefore, a certain degree of caution
is warranted in their use; namely, they should be employed in conjunc
tion with multiple criterion measures such as supervisor ratings. In
calling for multiple criterion measures of counseling effectiveness,
Goodstein and Grigg (1959) concluded that "client satisfaction is one
important factor in any multifactor approach to the problem of effec
tiveness in counseling" (p. 23).
Summary
Both supervisor and client ratings have been researched extensively
as effectiveness criteria. Supervisor ratings have enjoyed a high level

35
of acceptance and continue to be regarded as one of the most valid
criterion measures. Client ratings have yielded more mixed results, but
the weight of ev
ratings are used
is evidence that
dence supports their use. When supervisor and client
in conjunction, as they have been in this study, there
each strengthens the credibility of the other.

CHAPTER I I I
METHODOLOGY
This study attempted to investigate the extent to which effective
ness ratings of counselor trainees were related to preferred communica-
tional modes. Preferred verbal responses of counselor trainees were
examined in relationship to both client perceptions of counseling rela
tionships and supervisory evaluations of counseling effectiveness.
Description of the Sample
The subjects involved in this study included University of Florida
counselor trainees, the trainees' clients, and supervisors of the
trainees. A more detailed description of these samples, including the
process of selection, follows.
I. Counselor Trainees. The counselors-in-training group in this study
were composed of graduate students in the Department of Counselor
Education at the University of Florida. These counselor trainees
were all currently enrolled in a practical, supervised counseling
experience, referred to as a practicum or an internship, through the
Department of Counselor Education. Prior to this supervised train
ing period each trainee had completed an introductory course in the
helping process followed by a course which combined a survey of
counseling theories and approaches with a counseling laboratory.
The laboratory experience utilized role playing and video-taped
microcounseling (Ivey, 1971) with volunteer clients. Both of these
36

37
courses lasted for an academic quarter and preceded the first super
vised practicum. Students pursuing a specialist degree must com
plete three practica which typically amount to ten hours per week
and one internship quarter which is defined as a full-time, forty
hours per week counseling experience. For doctoral students the
total requirement is five practica and three quarters of intern
ship. Counselor trainees in this study were at any point in the
completion of their degree programs from the first practica to the
last internship. Trainees have available actual counseling oppor
tunities in a variety of academic and community settings. Counselor
trainees in this study were, for example, involved in a university
counseling center, a university mental health service, a community
mental health center, a drug treatment center, a delinquency treat
ment program, a paraprofessionaI training program, a community
college, a public school, a hospital, or any of a number of similar
human service agencies either affiliated with the University of
Florida or in nearby communities.
Graduate counseling students who were involved exclusively with
elementary and middle school children were not included in this
study since children that young often have difficulty completing the
inventories used in this study. All other counselor trainees
involved in a practicum or internship during the target quarter,
and who received supervision through the department, were asked to
participate in the data collection process.
2. Supervisors. The supervisors in this study were counselor educators
or professional counselors who had themselves completed rigorous
graduate training programs in counseling at the university level.

38
3.
All supervisors had extensive counseling experience in a variety of
settings with a wide range of clients. Most of these supervisors
had completed a counseling internship of a full academic year in
addition to other supervised practica or clerkships before entry
into full-orbed professional practice. Every supervisor was a
faculty member in the Department of Counselor Education or had been
approved by the department. Each supervisor held an earned doctorate
in an appropriate field such as Counselor Education or Psychology.
All of the supervisors had conducted weekly individual supervision
with the counselor trainee being evaluated for a period of not less
than six weeks.
Clients. The clients in this study were persons seen by the coun
selor trainee in a helping relationship at the trainee's assigned
agency during the predetermined week of data collection. Each
client had to have been seen by the counselor trainee subject for
a minimum of two sessions of at least one half hour duration to be
included in the study.
Data Col lection
Data collection for this study occurred during the seventh and
eighth weeks of the Spring, 1976, quarter at the University of Florida.
During that time frame each counselor trainee subject was evaluated by
an individual supervisor. Only counselor trainees who received super
vision on the university campus or in the local community were included
in the sample. Also, evaluative ratings were made by each trainee's
individual supervisor only. Generally, the individual supervisor met
weekly with the trainee on a one-to-one basis for one hour per week.

39
The focus of supervision was to review the trainees counseling efforts,
primarily via tape recordings and verbal report; to assess the trainee's
level of skill; and to guide the supervisees development toward full
professional competence. The method used to evaluate the effectiveness
of the supervisee was the Counselor Evaluation Rating Scale (see Appen
dix) which the supervisor returned to the researcher following comple
tion.
During the first week of data collection, the investigator arranged
with all relevant group supervisors to meet with their groups for the
purpose of administering the Helper Response Preference Inventory (see
Appendix) to each counselor trainee. At that time demographic data on
each trainee were also gathered by means of a questionnaire (see Appen
dix). The first two demographic variables, sex and major field, were
included to obtain some descriptive data on the counselor trainee sample.
They were omitted from the statistical analysis since the data were not
continuous. Altogether the following six sets of demographic informa
tion were requested:
1. Sex of the counselor trainee;
2. Undergraduate field of major;
3. Age of the counselor trainee;
4. Cumulative graduate grade point average at the end of the
previous quarters;
5. Number of graduate quarters completed at the end of the quarter
of data col lection;
6. Number of quarters of supervised counseling practice completed
at the end of the data collection quarter.
At each group supervision session the counselor trainees were

40
requested to administer the Barrett-Lennard Relationship Inventory to
one of their clients. The trainees were then informed about the adminis
tration procedure. Specific verbal instructions to the counselor
trainees were as follows:
From among the clients you have seen in a counseling relationship
for a minimum of two thirty-minute sessions, select your most
typical client. Ask the selected client to please complete the
Barrett-Lennard Relationship Inventory based upon your counseling
contacts together. Make sure the client reads the cover letter
of explanation. Also, be certain that the client thoroughly under
stands the rating procedure. Emphasize that he or she is the only
client you are asking so cooperation is extremely important.
Offer to stay with the client while the first page of the Rela
tionship Inventory is completed in private, to answer any questions
that might arise. When you are convinced the client understands
the task and will follow through, stress the fact that the instru
ment will be ready to mail as soon as completed and leave the client
to finish the ratings.
The client was then expected to comply with the verbal instructions
to evaluate and mail the instrument. Similar instructions were also
reinforced in the explanatory cover letter (see Appendix). The instru
ment was stamped and pre-addressed so the client could personally mail
the Inventory to the researcher. The brief cover letter, in an effort
to maximize the probability of genuinely candid responses, explained the
procedure in part. It was emphasized that even though the ratings were
important, they would not be used to personally evaluate the individual
counselor trainee. Also stressed was the fact that since all information

41
is strictly for research purposes the ratings would be analyzed as group
data only.
In order to identify missing data, each Relationship Inventory
carried the name of the counselor trainee being evaluated. Five days
following the period of data collection, the researcher began calling
counselor trainees to identify and rectify the causes of unreturned
Relationship Inventories. Postcards were mailed to counselor trainees
who could not be reached at home to inform them of the missing Inventory
and request followup. Thirty days after the original week of data
collection it was assumed that all completed Inventories had been
received. Statistical analysis of the available data then proceeded.
Whenever counselor trainees were absent from their group supervi
sion sessions, they were contacted by telephone to solicit their coopera
tion. When the HRPI was administered and the Relationship Inventory
delivered to these trainees, the same instructions were given as in the
group sessions. Also, the follow-up of missing data was the same as
previously described for trainees instructed individually.
Limitations
Since a deliberate sample of counselor trainees was selected for
study rather than a random sample from the entire universe of counselor
trainees, the genera I izabiIity is limited to graduate counselor educa
tion students at the University of Florida, their supervisors, and their
clients. Furthermore, the exclusion of elementary and middle school
counselors, although methodologically essential, further minimized the
external validity of the results obtained.
A second limitation of this study lies in the level of development

42
of the instruments employed as criteria of counseling effectiveness.
Neither the Barrett-Lennard Relationship Inventory nor the Counselor
Evaluation Rating Scale was validated directly upon client outcome
research. Therefore, the measurement of counseling effectiveness was
indirect, but, nevertheless, consistent with procedures generally
recognized as appropriate within the professional field of counseling.
One should not, however, construe the findings of this study to be
equivalent to validation based upon actual outcome measures of client
change. Likewise, the results of this study are specific to the instru
ments used and should not be generalized to other comparable instruments
until further study demonstrates which global measures of counseling
effectiveness may be considered statistically interchangeable.
Instrumentation
The three instruments referred to in the presentation of the metho
dology are each examined independently in the following section.
Helper Response Preference Inventory (HRPI)
The Helper Response Preference Inventory (HRPI) was originally
devised by Dr. James L. Lister in the early 1970's in an extension of
the work of Porter (1950) on classification of counseling responses.
The HRPI (see Appendix) consists of fifty items, each presenting a
statement by a he I pee to which the respondent must choose one of two
optional helper statements as a response. The intent of the forced-
choice format was to pair twenty responses, of approximately equal
valence, from each of the five following categories: understanding,
supportive, interpretive, probing, and evaluative responses. Thus, the
one hundred total responses enabled each set of possible pair-choices

43
such as understanding-probing, or interpretive-evaluative to occur five
times each. The similar pairs were randomly distributed throughout the
test items. Theoretically, this method was believed to allow identifi
cation of a counselors preferred response style by quantifying the
over-chosen response categories.
Validity studies on the HRPI were conducted by Bradley (1973) in
an unpublished study of both counselor trainees and a paraprofessionaI
group of child care workers. Using the Counselor Evaluation Rating
Scale as a criterion, the regression studies for a sample of 65 counselor
trainees revealed that the HRPI scores and ratings on the CERS were
correlated .47 overall. The counseling scale of the CERS alone corre
lated .48 with HRPI scores.
In the companion investigation, HRPI scores were correlated with a
global subjective rating of child care workers by their immediate super
visors as effective, adequate, or relatively ineffective helpers. A
multiple correlation of .63 was yielded, but the understanding score
alone correlated .61 with supervisor rankings.
Test-retest reliability of the HRPI was found to be .83 for a
heterogeneous sample of 43 subjects after a span of four to six weeks
(Bradley, 1973).
Split-half reliabilities were computed by Lister (1972) for the
five original sub-scales as well as a composite scale based upon forty
consensus items agreed upon by a sample of eleven doctoral level coun
selor educators on the faculty of the University of Florida. This con
sensus scale was labeled the "expert" scale. The split-half reliabili
ties for the a priori scales and the "expert" scale were generally
modest and are summarized in Table I for three separate sample groups.

44
TABLE I
SPLIT-HALF RELIABILITIES FOR THREE SAMPLES: GRADUATE LEVEL COUNSELOR
TRAINEES, PARAPROFESSIONAL COUNSELOR TRAINEES IN A COMMUNITY COLLEGE,
AND PARAPROFESSIONAL DISASTER RELIEF COUNSELOR TRAINEES
Santa Fe Community
College Human Service
Aide Trainees
(N=6I)
Counselor Education
Graduate Students
(N-72)
Wilkes-Barre Flood
ParaprofessionaI
Emergency Counselor
T rainees
(N=50)
Expert
.62
.62
.49
Understanding
.53
.52
.52
Supportive
.41
.47
.42
1nterpretive
. 17
.42
.34
Probing
.18
.30
.52
Evaluative
.13
.37
.29
Factor analysis of the HRPI was conducted in 1973 (Bradley) pooling
319 sets of responses in the sample. Since the items were forced
dichotomies, it was imperative to compute tetrachoric correlations. The
items were factor analyzed, and the principal axis solution rotated to
varimax criterion (Guertin & Bailey, 1970). Highest coefficients
served as communality estimates. Two items had to be deleted because
of low cell distributions. The remaining 48 item variables were factor
analyzed and again rotated to varimax criterion, but a clear factor
structure did not emerge until the principal axes factor matrices were
rotated in an oblique rotation procedure. Ten principal axes factors
were chosen because the latent roots tended to plateau beyond that point
and all ten factors chosen had latent roots greater than 1.0. The

45
oblique factor matrix was clear with ten factors accounting for 60.2%
of the total score variance and 87.6% of the common variance. Factors
were limited to items that had at least a .30 factor loading. Factor
scores were obtained then by weighting each item in proportion to its
factor loading as follows:
Factor Scores
Factor Loading Range Weight of the Item
.30 to .39 I
.40 to .49 2
.50 to .59 3
.60 to .69 4
.70 to .79 5
The relative independence of the oblique factor structure is
attested to by the low correlations between primary factors. All inter
correlations are below .30 except between factors A and J, which is
0.38.
The 48 items, which were the variables factor analyzed, presented
alternative choices of responses perceived as being helpful. In each
item a value judgment was required by the respondent of which item to
choose and which to reject. Consequently, every response represented
a preference of one response over another one. The factors corres
pondingly have a high end which represents the responses chosen and a
low end which represents the responses rejected. The factor labels,
which are believed to describe the common theme of the items loading in
each factor group, are listed below. Also presented are the number of
items loading the factor and the total weights for the factor. Factors
are listed in descending order according to the proportion of total
score variance accounted for by the factor.

46
TABLE 2
FACTOR LABELS AND FACTOR SCORES FOR THE HELPER
RESPONSE PREFERENCE INVENTORY
Preferred Response
Factor Label
Rejected Response
Factor Label
Number
of Items
Total (
Weight
A.
Empathic Under
stand i ng
Probing
12
27
B.
Confrontation
False Reassurance
9
21
C.
Specificity
Ambiguity
8
20
D.
Va 1idation of
Experience
Analytical Interest
6
12
E.
Clarificat ion
Superficia1ity
9
16
F.
Focused Probing
Premature Judgment
5
1 1
G.
Tentative Formu-
1 at ions
Avoidance of Emotion
4
7
H.
Requesting In
formation
Giving Information
4
8
1.
Helpee's Intern
al Reference
Helpers External
Reference
5
8
J.
Frank Feedback
1 ncongruence
5
10
Barrett-Lennard Relationship Inventory (BLRI)
The theoretical rationale which led to the construction of the
Barrett-Lennard Relationship Inventory (Barrett-Lennard, 1962) was
derived from the formulation by Carl Rogers (1957) of "necessary and
sufficient conditions" for beneficial client change. The Relationship
Inventory was intended to assess four aspects of any two-person rela
tionship, particularly the counselor-client relationship. Those four
aspects were empathic understanding, level of regard, congruence, and

47
unconditionality of regard. The original instrument contained 85 items
but was soon shortened to 72 items, tost of the research on the Rela
tionship Inventory before mid-1964 utilized the 72-item "interim revi
sion" of the original investigative instrument. Since that time, the
64-item revision, or adaptations of this revision, has been employed
in most research using the Inventory. The author (1969) explains the
procedures used in revising the instrument to its current form:
The 64-item revision was based (a) on item-analysis results
from several samples of data obtained with earlier versions
of the instrument,(b) on other technical considerations,
such as achieving a numerical balance of positively and
negatively stated items in each of the four scales, (c) on
minor theoretical refinements, particularly in respect to
the unconditionality of regard scale and (d) on the con
cern to alter or replace items of a relatively abstract
or difficult kind. The item analysis provided empirical
checks on what each previously used item was doing and
gave results that were used in conjunction with b), c)
and d) to refine the selection and preparation of items.
In general, the aim of the revision was to further im
prove the sensitivity and versatility of the instrument
for measuring the defined variables in the context of any
significant interpersonal relationship. (Barrett-Lennard,
p. I, 1969)
Well over one hundred research studies have used the Barrett-
Lennard Relationship Inventory establishing the instrument as one
of the most widely used measures of the quality of human relation
ship available today. In its current form, the Relationship Inven
tory's 64 items include 16 for each of the four subtest dimensions,
which assess a person's perception of another person in a relation
ship. Each statement is scored on a 6-point scale, from +3, "yes,
I strongly feel that is not true." There is no 0 response. The
items are general enough to assess practically any two-person rela
tionship which the examiner may specify.
The published reliability studies on the earlier 72-item form

48
of the Relationship Inventory were acceptable and generally consistent
across investigations. Barrett-Lennard (1962) reported split-half
reliability coefficients for the various relationship dimensions within
a range of .82 to .93. In two separate studies, Snelbecker (1961,
1967) discovered a range of split-half reliabilities between .75 and
.94. The range was .83 to .95 for Hollenbeck (1965) and .82 to .91
for Hough (1965).
The test-retest reliability coefficients reported by Barrett-
Lennard (1962) after a four-week interval, ranged from .84 to .90. After
a six-months interim between administrations Hollenbeck (1965) found a
somewhat lower range of reliability, on the 72-item form, of from .61
to .81. In an investigation using the group form of the Relationship
Inventory, Berzon (1964) reported a reliability coefficient of .86 for
total scores after a four-week interval.
Refinement procedures have helped the reliability coefficients for
the 64-item revision of the Relationship Inventory to run as high or
higher than with previous versions of the instrument. Several test-
retest studies are summarized in Table 3.
The question of the validity of the Barrett-Lennard Relationship
Inventory is interwoven with the complexities of obtaining direct cri
terion-based validity checks in the overall field of counseling.
Barrett-Lennard (1969) relates this difficulty to "the absence of alter
native, established measures of theoretically equivalent dimensions of
perceived interpersonal response" (p. 4). Although a number of
researchers have been basically in agreement with the statement by Mills
and Zytowski (1967) that "the test has, at a minimum, face validity,"
(p. 194) a more persuasive justification for its use as a criterion
instrument is essential.

49
TABLE 3
TEST-RETEST RELIABILITY OF THE SUBTESTS OF THE
64-ITEM RELATIONSHIP INVENTORY
IN SEVERAL STUDIES
Relationship
Dimensions
Barrett-Lennard
(1969)
Form OS,
N=40
2 to 6 weeks
Barrett-Lennard
(1969)
Form MO,
N=38
12 days
Mills and
Zytowski
(1967)
Form MO,
N=79
3 weeks
Mills and
Zytowski
(1967)
Form OS,
N=79
3 weeks
Level of
Regard
.88
.79
.86
.74
Empathic
Understanding
.86
.91
.84
.90
Unconditiona-
lity of Regard
.86
.86
.80
.80
Congruence
.92
.85
.87
.88
Tota 1 Score
.92
.89
It must be recognized that evidence supporting the validity of the
Barrett-Lennard Relationship Inventory is cumulative rather than con
clusive. The bulk of evidence derives from research, much of which was
carefully designed, in which the associations between Relationship
Inventory measures and other variables were predicted on the basis of
logical and theoretical grounds. When these findings are positive, the
underlying theory gains credibility; and the argument that the dimen
sions, purported to be measured, are indeed related to measurable rela
tionship variables of fundamental importance in the consequences of
human interaction becomes more believable. The assumption, then, was
that a mounting body of findings which were consistent with the constructs

50
upon which the instrument was based established a defensible justifica
tion for utilizing the Relationship Inventory as a criterion in this
study.
The basic theory underlying the Relationship Inventory (Rogers,
1957; Barrett-Lennard, 1962) has been substantially supported by the
vast range of studies using the Truax and Carkhuff rating scales to
measure levels of therapeutic conditions within the context of relation
ship (Truax & Carkhuff, 1964, 1967; Carkhuff & Truax, 1965a, 1965b;
Berenson, Carkhuff, & Myrus, 1965).
More pointed research has involved the instrument per se. Success
ful use of the Relationship Inventory to appraise counselor-client rela
tionships has been reported by Van der Veer (1961), Barrett-Lennard
(1962), Cahoon (1962), and Gross and Deridder (1966). Other intimate
relationships measured with significant results include marital partners
(Thornton, I960), student-parent (Hollenbeck, 1965), supervisor-teacher
(Blumberg, 1968), and student-teacher (Mason and Blumberg, 1969).
Barrington (1961) related scores on the Barrett-Lennard Relation
ship Inventory to such specific counselor behaviors as a decrease in
the length of counselor responses from the first to the second inter
view, percentage of emotional words used by the counselor, and simi
larity of the counselor's rate of speech to the client's rate. The most
predictive sub-scale was the empathic understanding dimension, but
perceived counselor congruence was highly related to more integrated
verbalizing by the counselor in the second hour of counseling.
Using a Q-sort technique, Emmerling (1961) studied the relation
ship of "openness" in relatively effective and ineffective teachers to
how their students perceived them in relationship. It was hypothesized

51
that "open" teachers would see themselves as responsible for diffi
culties and remedial action in their classroom work. The rationale was
that such teachers would tend to be more genuine, sensitive, and
accepting than teachers who, under similar conditions, see problems as
unrelated to their own characteristies and externally imposed. The
finding that students of the more "open" teachers did perceive them more
positively on each of the Relationship Inventory dimensions may be
interpreted as implying that the measures discriminated differences
consistent with theoretical predictions.
Cahoon (1962), in another well-known investigation, found signifi
cant relationships between the process experiential levels and open-
mindedness, on one hand, and the client-perceived quality of the coun
seling relationships on the other. For his sample of practicum-Ievel
counselor trainees he employed the Process Scale, the Dogmatism Scale,
and the Barrett-Lennard Relationship Inventory.
The uses of the Relationship Inventory have been extensive.
Donnan, Harlan, and Thompson (1969) used the original 85-item form to
significantly correlate four personality factors on the 16 PF with
three relationship variables for a sample of 22 counselors who were
evaluated by 880 prospective college freshmen following counseling. In
another study, Brown and Calis (1968) measured the effect upon the
clients perception of the relationship with his or her group counselor,
depending upon whether the group was required or self-initiated. Lanning
(1971) used the Relationship Inventory to evaluate whether practicum
students perceive the supervisory relationship the same in group and
individual supervision. A still different application of the Inventory
was made by Armstrong (1969) who examined the difference between students

52
in counseling and students not in counseling with regard to quality of
their relationships with intimate friends and the tendency to use
intimate friends as therapeutic agents. In a very recent study Loesch
(1975) used the Relationship Inventory to assess whether a client's
nonverbaIized feelings and the counselors' perception of them during
the interview were related to subsequent counselor and client percep
tions of the counseling relationship.
The appropriateness of using the separate relationship dimensions,
purported to be measured by the Relationship Inventory, rather than the
more global total score, has been the subject of some controversy.
Initially Barrett-Lennard (1962), in asserting that separate relation
ship aspects were indeed being measured, stated that the scales were
not measuring a general factor, such as the client's overall satisfac
tion or dissatisfaction with the therapy relationship. However, Mills
and Zytowski (1967), on the basis of a princi pa I-components analysis
of the intercorrelations between the subtests for several samples using
a variety of forms of the Inventory, concluded differently. They found
that a general component accounted for about two-thirds of the variance.
WaIker and Little (I 969) chalenged the findings of Mills and
Zytowski since their conclusions were based upon an unrotated principal-
components analysis. Walker and Little then combined the Inventory
scores of 150 university undergraduates and conducted a component
analysis of the 64 items. They found three factors which they labeled
"non-evaIuative acceptance," "psychological insight," and "I ikabiIity."
At about the same time Tosi, Frumkin, and Wilson (1968) did inter-
correlational studies on the subtests of the Relationship Inventory
scores of 69 clients at the Kent State University Guidance Laboratory.

53
They concluded, from the correlation range between subscales of from
.31 to .58, that their "data appear to support Barrett-Lennard's con
clusion that the four subscales are relatively independent" (p. 642).
In 1974 Lanning and Lemons administered the OS form of the Rela
tionship Inventory on three separate occasions to compile data for a
pri ncipa I-components analysis and an orthogonal rotation of the factor
matrix by the varimax procedure. They found two nearly orthogonal
factors with Level of Regard, Empathy, and Congruence loading on a
component that accounts for from 62 to 72 percent of the total variance.
Unconditionality of Regard loads predominantly on the other factor which
accounts for 16 to 21 percent of the total variance. The authors felt
their study provided some support for the earlier findings of Mills and
Zytowski. Lanning and Lemons concluded:
The major component is probably a measure of the overall
satisfaction with the relationship, contrary to earlier claims
made about the separateness of the subtests. This indicates
that there may be one predominant characteristic across all
of the relationships measured rather than four separate ones.
(1974, p. 230)
Since there appeared to be somewhat inconclusive empirical support
for or against a general relationship satisfaction factor, this study
employed the total score of the Barrett-Lennard Relationship Inventory
and each sub-score as criterion measures based upon client ratings of
perceived counseling effectiveness. Both the male and female forms of
the OS-64 revision (see Appendix) were utilized in this study.
Counselor Evaluation Rating Scale (CERS)
The Counselor Evaluation Rating Scale (see Appendix) which was
used in this study as a criterion measure of supervisor rating, com
prises 27 items that enable a supervisor to rate a counselor trainee's

54
performance in a supervised practicum (Myrick & Kelly, 1971). The items
are designed to assess the counselor trainee's understanding of a
counseling rationale, counseling practice with counselees, and explora
tion of self and counseling relationships with a supervisor. A 7-point
(+3 to -3) Ukert-type scale is employed to mark each item. The CERS
yields three scores: a counseling subscale (C) is an appraisal of the
counselor-trainee's individual counseling behavior; a supervision sub
scale (S) is an assessment of the person's involvement and progress in
the supervisory process; and a total score (T) of these two subscales
plus one additional item: "Can be recommended for a counseling posi
tion without reservation." The resulting total score ranges between 27
and 189 when the negative scores are eliminated. The counseling scale
and the supervision scale each yield scores between 13 and 91.
In their original article (1971) the authors, Myrick and Kelly,
reported a reliability study involving 45 student counselors and their
supervisors from the University of Florida. Individual supervisees,
who met with their assigned faculty supervisor for a minimum of one
hour per week for twelve v/eeks, were rated by these supervisors on the
CERS at the end of a fall quarter practicum. Utilizing the Spearman-
Brown correction of the split-half reliability procedure, the investi
gators found a coefficient of .95. When the 13 counseling items were
compared with the 13 supervisor items, a correlation coefficient of .86
was produced.
The stability of the CERS over a period of time was determined by
a test-retest reliability procedure. After a minimum of a four week
lapse in time since collecting the CERS ratings, the supervisors were
again requested to rate the same student counselors with whom they had

55
been involved during the fall quarter. Data analysis yielded a .94
product moment reliability coefficient.
Myrick and Kelly (1971), in their presentation of the instrument,
reported they had failed to find relationships between either Graduate
Record Examination scores or undergraduate grade point averages and CERS
ratings. The correlation coefficients were respectively -.03 and -.25.
Wittmer and Lister (1971) also correlated Graduate Record Examina
tion scores and CERS ratings for 53 counselors in training whose CERS
ratings were compiled following the first quarter of practicum. The
GRE had been taken prior to admission. The correlation of the GRE
verbal score and CERS ratings was -.08. When the GRE quantitative score
was correlated with the CERS, the coefficient yielded was .12. Neither
correlation was significant at the .05 level. However, this same study
also compared results of each counselor trainees 16 PF profile, taken
the quarter before the first practicum, with his CERS ratings. The
authors employed the index of predicted counseling effectiveness which
is a regression equation based upon studies of the 16 PF at the Insti
tute for Personality and Ability Testing (1963). The Pearson product-
moment correlation coefficient produced was .41, which was significant
beyond the .01 level.
In another published article using the CERS as a criterion of
counseling effectiveness, Myrick, Kelly, and Wittmer (1972) found that
scores correlated significantly with certain factors of the 16 PF. The
investigators defined 20 effective counselor trainees with high CERS
scores and 20 ineffective counselor trainees with low CERS scores. The
15 trainees with scores near the mean of the CERS were eliminated. An
examination of the 16 PF scores of the two comparison groups yielded

56
eight personality factors with score differences between the groups
which were statistically significant.
In a challenge>of the previous research, Jones (1974) questioned
the appropriateness of using an instrument designed to help supervisors
evaluate practicum students behavior as a criterion of counseling
effectiveness. He argued, "To be valid as a criterion measure of coun
selor effectiveness, an instrument such as the CERS must be related to
counseling outcome or some other aspect of job performance" (p. 113).
Following his line of reasoning, Jones attempted to investigate the
relationships between the CERS and empathy, genuineness, and respect
using the Carkhuff Scales since these variables "have been established
as being significantly related to counseling outcome" (p. 113). He
found a significant positive relationship only between the level of
respect offered in the counseling situation and supervisor ratings on
the S (supervision) scale of the CERS. He concluded:
. . there appears to be no relationship between levels
of empathy, respect, and genuineness and the other CERS
scales, although the positive correlations between respect
and the CERS scales of C (counseling) and T (total) may
be noteworthy. (p. 115)
Forster and Hamburg (1976) in a recent study of 30 full-time coun
selor trainees failed to find a significant relationship between CERS
ratings and 16 PF scores.
A very recent study by Loesch and Rucker (1976), which has been
submitted for publication, has provided some support for the original
CERS Scales. In a second order factor analysis of more than four
hundred sets of ratings they obtained an oblique solution which closely
approximates the original classification of items into supervisory and
counsel ing sea Ies.

57
In the report on validity studies using the Helper Response Prefer
ence Inventory, it was pointed out that for a sample of 65 counselor

trainees there was a .47 correlation between HRPI scores and overall
CERS scores. The correlation was .48 for the counseling scale of the
CERS when related to the HRPI scores separately.
Borman and Ramirez (1975) compared the CERS self ratings of 25
master's degree counseling students in practicum with their ratings by
their supervisor and practicum assistants who were doctoral students.
Their findings revealed that there were significant differences found on
9 of the 27 items. However, most of the differences were attributed to
the students rating themselves higher than supervisors or assistants
and, specifically, differences in ratings between students and practicum
assistants. On only one item, dealing with the degree of relaxation and
comfort in the supervisory sessions, was there a statistically signifi
cant difference between the supervisor's ratings and the counselors'
self-ratings.
Analysis of Data
The data collection process, previously described, yielded four
basic groupings of data: demographic variables, Helper Response Pre
ference Inventory scores, Barrett-Lennard Relationship Inventory scores,
and Counselor Evaluation Rating Scale scores. Using the two criterion
instruments eight dependent criterion variables were isolated: the
Relationship Inventory total score plus the four subscores, and the
counseling, supervisory, and total scales from the Counselor Evaluation
Rating Scale. By means of the SPSS library computer program for stepwise
regression (1973), the ten factor scores of the Helper Response

58
Preference Inventory and the four demographic factors were predictor
variables regressed upon each of the dependent criterion variables. The
stepwise solution program (Kerlinger, 1973) allows at each step the
variable that makes'the greatest increment to the multiple coefficient
2
of determination, R to enter the computational process and calculates
a regression equation for that step. Also at each step of the analysis,
several calculations are performed and printed. These include the mul
tiple correlation coefficient, the regression equation, partial correla
tions, and the regression and residual sums of squares. One advantage
of this stepwise program is that the investigator can specify the level
of F ratio for entering a variable into the equation and the level of
F for removing a variable from the equation. For this study an alpha
level of .05 was considered the level of significance necessary to fail
to reject the hypotheses.
Utilizing the stepwise regression procedure made it possible to
ascertain which combination of independent variables, HRPI scores,
demographic data or combinations thereof, was the most efficacious pre
dictor of the dependent criterion variables, the Relationship Inventory
or the Counselor Evaluation Rating Scale. The regression equation
computed specified the precise beta weights for the independent variables
which, taken in combination, yielded the greatest reduction in the error
sums of squares when predicting either of the criterion variables.
2
This method of analysis also permitted an examination of the R s
and F ratios produced by each variables increase in the overall pre
diction process. By observing the increment to prediction for each
variable, the relative efficacy of different variables in the regression
equation became apparent. In effect, each hypothesis was tested for the

59
optimum combination of predictor variables as well as for any separate
predictor variable which was, by itself, a significant predictor of
either of the dependent variables.
While not central to the purposes of this study, canonical corre
lations between all ten HRPI variables and the four Barrett-Lennard
Relationship Inventory subscore variables were computed to establish
the maximum correlation possible between the two sets of variables and
to estimate the variance shared by the two composites of variables. The
same procedure was used with the three Counselor Evaluation Rating
Scale scores and the ten HRPI variables. Finally, canonical correla
tions were computed for the four Barrett-Lennard Relationship subscore
variables and the three Counselor Evaluation Rating Scale score varia
bles. An intercorrelation matrix of all the variables in this study
is presented on page 67. It was anticipated that an intercorrelation
matrix could help clarify the results and could possibly explain the
presence or absence of statistically significant relationships among the
variables under scrutiny.

CHAPTER IV
RESULTS OF THE STUDY
Introduction
The research procedures for this study were enacted according to
the experimental design described in Chapter III. Analysis of the data
was conducted via the computer facilities of the Northeast Regional Data
Center under the auspices of the Florida State University System. Spe
cifically, the stepwise regression subprogram and the canonical corre
lation subprogram of the Statistical Package for the Social Sciences
developed by Nie et al. (1975) were the analytical procedures utilized.
The results of these statistical operations are presented below in a
tabular format. This chapter consists of: (I) a descriptive profile
of the counselor trainee sample, including demographic data on the
subjects, and the scores of all subjects on the two criterion instru
ments as well as the predictor instrument; and (2) the statistical tests
of the two major hypotheses and eight minor hypotheses subsumed under
the two major ones, presented in sequential order.
Descriptive Data of the Sample
The Counselor Trainees
There were 67 counselor trainees who appeared to meet the criteria
for inclusion: namely, having clients on a continuing basis who
possessed the literacy skills requisite to valid completion of the
60

61
Relationship Inventory. Of those 67 counselor trainees, two chose not
to participate for personal reasons, one was unavailable during the data
collection period, and three failed to return even the demographic data
sheet and HRPI answer sheet after agreeing to do so, despite follow
up attempts. At least partial data were compiled on 61 counselor
trainees. Supervisors' ratings were obtained on all 61 subjects in
addition to the demographic data and the HRPI answer form. However,
the client rating on the Relationship Inventory was returned on only
46 counselor trainees. Therefore, the net sample for this study was
46 counselor trainees.
Descriptively, the counselor trainee sample was composed of 27
males and 19 females. The age range was 22 to 51 years, and the mean
age was 27.6. The graduate grade point average ranged from 3.20 to 4.00
with a mean of 3.77. The counselor trainees in the study had completed
from 3 to 17 graduate quarters and I to 8 quarters of supervised prac
tice with respective means of 7.6 and 4.1. A breakdown of demographic
data on the counselor trainees is presented in Table 4.
The counselor trainees came from a wide variety of undergraduate
major backgrounds. Although the majority of trainees (27) had chosen
psychology as an undergraduate major, eight additional fields were
represented. There were five trainees each from education and language
undergraduate programs. Likewise sociology, religion, and the physical
sciences were represented by two counselor trainees each. There was
one counselor trainee each from political science, art history, and
business. The overall impression is that the sample represents a
diversity of backgrounds, interests, and personalities.

62
TABLE 4
DEMOGRAPHIC DATA ON THE 46 PARTICIPATING
COUNSELOR TRAINEE SUBJECTS
Variable
Age
Mean
27.57
Standard deviation
5.24
Median
26.00
Range
22-51
Sex
Ma 1 e
27
(58.7$)
Female
19
(41.3$)
Grade Point Average
Mean
3.77
Standard deviation
.20
Range
3.20-4.00
Number of Graduate Quarters Completed
Mean
7.63
Standard deviation
3.26
Range
3-17
Number of Quarters of Supervised
Counseling Comp
1 eted
Mean
4. 10
Standard deviation
2.26
Range
1-8
Undergraduate Major Field
Psychology
27
(58.7$)
Education
5
(10.9$)
Language
5
(10.9$)
Sociology
2
( 4.3$)
Physical Sciences
2
( 4.3$)
Re 1 igi on
2
( 4.3$)
Political Science
1
( 2.2$)
Art History
1
( 2.2$)
Business
1
( 2.2$)

63
Supervisors
The supervisors in this study numbered 27. There were 20 males and
7 females. With the exception of one psychiatrist all of the super
visors held an earned doctorate in counseling, guidance, student personne
work, or some allied branch of psychology. Two supervisors gave
individual supervision to four trainees, and therefore each rated four
counselor trainees. Three supervisors completed ratings on the three
trainees they supervised individually, and seven supervisors rated two
trainees. The remaining fifteen supervisors each contributed a single
set of ratings on one counselor trainee. Only two female supervisors
completed more than one trainee's ratings, and in both cases the number
of ratings was two.
Clients
To intentionally preserve a sense of anonymity there was no attempt
to gather descriptive information on the client sample. While such knowl
edge would of course be interesting, there would be no way to compare
the clients who responded with those who for one reason or another never
returned the Relationship Inventory. Consequently, client data would be
of limited interpretive value.
Client Evaluation Scores on the Counselor Trainees
Converted scores from the evaluative raw data are presented in
composite form in the following two tables. The client ratings on the
Barrett-Lennard Relationship Inventory are collapsed into means,
standard deviations, and ranges in Table 5. The scores are character
ized by broad ranges with 46 to 47 points of spread among the four sub
scales. The total score range is 182 points.

64
TABLE 5
MEANS, STANDARD DEVIATIONS, AND RANGES OF SCORES ON THE
BARRETT-LENNARD RELATIONSHIP INVENTORY (N=46)
Relationship
Dimensions
Means
Standard
Deviations
Range
Level of Regard
35.83
10.27
2-48
Empathic Understanding
30.54
12.84
1-48
Unconditionality of Regard
21.04
12.02
1-48
Congruence
33.52
11.50
1-48
Total Score
118.35
42.91
10-192
Of interest is the fact that the means of the sub scales all fall
in the low to mid thirties except for the Unconditionality of Regard
score which dips to 21.04. Nevertheless, there is no appreciable
difference in the standard deviations of all four subscores which fall
in the range of 10.27 to 12.84. The Total score had a mean of 118.35
and a standard deviation of 42.91.
Supervisor Evaluation Scores on the Counselor Trainees
The individual supervisor ratings on the Counselor Evaluation
Rating Scale are presented in Table 6. The summary shows the means,
standard deviations, and ranges of the scores for all 46 counselor
trainee subjects. Ranges on both the Counseling and Supervision sub
scales are quite broad, 33 to 91 and 38 to 91, respectively. The Total
score reflects a 91 to 189 point range.
The means of the two subscales, 81.33 for Counseling and 84.22 for

65
TABLE 6
MEANS, STANDARD DEVIATIONS, AND RANGES OF
SCORES ON THE COUNSELOR EVALUATION
RATING SCALE (N=46)
CERS
Sea 1es
Means
Standard
Deviations
Ranqe
Counseling
rO
fO
oo
10.35
33-91
Supervision
84.22
9.98
38-91
Total
171.54
20.05
91-189
Supervision, are comparable. Both mean scores fall slightly less than
one standard deviation below the maximum possible score, a phenomenon
observed as well for the 171.54 mean of the Total scores.
Scores of the Counselor Trainees on the Predictor Instrument
The means, standard deviations, and ranges of scores on the ten
factor scales of the Helper Response Preference Inventory are presented
in Table 7. It should be kept in mind that each factor represents a
chosen direction of responses and a rejected direction of responses.
Higher scores reflect a preference for the response group named first in
the parenthetical pair, and lower scores indicate that the response
group placed second in the parenthesis was preferred. All of the means
fall in the upper half of possible score range for the factor except
for Factor D (Analytical Interest Validation) where the preference
favored slightly the validation side of the dichotomous scale. For
four factors the entire score range observed fell in or very close to

66
TABLE 7
MEANS, STANDARD DEVIATIONS, AND RANGES OF SCORES ON THE
HELPER RESPONSE PREFERENCE INVENTORY (N=46)
Factors
Means
Deviations
Ranqe
A.
(Empathic Understanding-
Probing)
24.22
4. 1 1
17-27
B.
(Confrontation-False Reas
surance)
16.26
3.21
8-21
C.
(Specificity-Ambiguity)
18.11
2.68
10-20
D.
(Analytical 1nterest-Va1i-
dation)
4.13
2.60
0-10
E.
(Clarificat ion-Superficia1ity)
13.57
2.35
8-16
F.
(Focused Probing-Premature
Judgements)
8.02
2.68
l-l 1
G.
(Tentative Formulations-
Avoidance of Emotion)
5.76
1.75
2-7
H.
(Requesting-Giv¡ng Information)
4.61
1.48
3-8
1.
(1nterna1-Externa 1 Frame of
Reference)
5.17
2.03
0-8
J.
(Frank Feedback-1ncongruence)
8.52
1.55
5-10
the upper half of possible scores. Three factors (D, F, and I) yielded
scores that covered almost the entire possible range. Factors B, G,
and H produced scores somewhere between these two extreme groups. No
other discernible pattern appeared to emerge from this set of data.
Relationships Among the Three Instruments
Table 8 presents the intercorrelation matrix for the two criterion

TABLE 8
INTERCORRELATION MATRIX FOR ALL VARIABLES OF THE BARRETT-LENNARD RELATIONSHIP
INVENTORY, THE COUNSELOR EVALUATION RATING SCALE, AND
THE HELPER RESPONSE PREFERENCE INVENTORY
BLRI
BLRI
BLRI
BLRI
BLRI
CERS
Tota 1
Regard
Empathy
Unconditiona1ity
Congruence
Tota 1
BLRI Total 1.00
0.84
0.93
0.67
0.90
-0.26
BLRI Regard
1.00
0.74
0.47
0.73
-0.05
BLRI Empathy
1.00
0.47
0.84
-0.28
BLR 1 Unconditiona1ity
1 .00
0.40
-0.31
BLRI Congruence
1.00
-0.23
CERS Tota 1
1 .00
CERS Counsel¡ng
CERS Supervision
HRPI A
HRPI B
HRP I C
HRP I D
HRP I E
HRP I F
HRP I G
HRP I H
HRP I I
HRP I J
CTi

Table 8Continued
CERS
CERS
Counsel
Supervisi on
HRPI A
HRPI B
HRPI C
BLR 1
Total
-0. 16
-0.31
0. 16
0.02
0.00
BLRI
Regard
0.02
-0. 1 1
0.25
0.01
-0.01
BLRI
Empathy
-0. 18
-0.33
0.07
-0.02
-0. 12
BLRI
Unconditiona1ity
-0.23
-0.34
0. 12
-0.06
0. 1 1
BLRI
Congruence
-0. 15
-0.27
0. 19
0. 1 1
0.05
CERS
Tota 1
0.92
0.92
0.08
0. 18
0.23
CERS
Counse1ing
1.00
0.71
0. 16
0.19
0.19
CERS
Supervision
1 .00
-0.03
0.13
0.21
HRPI
A
1.00
0.24
0.28
HRPI
B
1.00
0. 19
HRPI
C
1.00
HRPI D
HRPI E
HRP I F
HRPI G
HRPI H
HRPI I
HRPI J
HRPI D
0.06
0.00
0.05
0. I I
0.03
-0. 14
-0.22
-0.03
-0.35
0.08
-0.20
I .00
oo

Table 8Continued
HRPI E
HRPI F
HRPI G
HRPI H
HRPI 1
HRPI J
BLRI
Tota 1
0.05
-0.20
-0.01
-0.20
-0.03
0. 1 9
BLRI
Regard
0. 12
-0. 1 1
-0.01
-0.24
-0.06
0.41
BLRI
Empathy
0.03
-0.21
-0. 1 1
-0.24
-0.06
0.17
BLRI
Unconditiona1Ity
-0.01
-0. 1 1
0. 17
-0.03
-0.05
-0.13
BLRI
Congruence
0.09
-0. 17
-0.07
-0.16
0.06
0.25
CERS
Tota 1
0.27
0.04
0.01
-0.05
0.08
0.13
CERS
Counsel 1ng
0.32
0.00
-0.01
-0. 1 1
0.05
0. 10
CERS
Superv1sion
0. 18
0.08
0.00
0.02
0.10
0. 14
HRP 1
A
0.50
0. 1 1
0. 19
-0.39
0.08
-0.08
HRP 1
B
0.26
0.17
0.12
-0.16
0.41
0. 16
HRP 1
C
0.20
-0.00
0.50
-0.03
0.14
0.03
HRP 1
D
-0.02
0. 17
-0.14
0.05
-0. 19
0.14
HRP 1
E
1.00
0. 13
0.25
-0.20
0. 1 1
0.08
HRP 1
F
1 .00
-0.09
-0.06
0.22
-0.09
HRP 1
G
1.00
-0.04
0.01
0.06
HRP 1
H
1 .00
0. 10
0.00
HRPI
1
1.00
0. 16
HRPI J 1.00
as
\o

70
instruments, the Barrett-Lennard Relationship Inventory (BLRI) and the
Counselor Evaluation Rating Scale (CERS), and for the predictor instru
ment, the Helper Response Preference Inventory (HRPI). There are 204
positive correlations and 120 negative correlations.
Significant in this matrix are the consistently negative correla
tions between BLRI scores and CERS scores. The mean correlation for all
correlations between the two criterion instruments is -.21. All correla
tions are negative except for the correlation between the BLRI Level of
Regard score and the CERS Supervision score which is a positive, though
ngligible, .02.
Canonical correlations were computed between subscale scores of the
BLRI and subscale scores of the CERS. Similar operations were performed
relating the HRPI variables to each set of the criterion instrument
subscales. In all cases the canonical correlations failed to reach the
.05 level of significance.
The relationships of each of the ten HRPI factors to the combined
scores of the BLRI and also to the combined scores of the CERS are
presented in Table 9. The statistic was derived by averaging the
correlations of the predictor HRPI factor with all subscales and total
score of each criterion instrument. The first column reveals the mean
correlation of the specified HRPI factor with the five subscores of the
BLRI. The second column shows the mean correlation of the specified
HRPI factor with the three subscores of the CERS. Notable are the seven
correlations, either positive or negative, above .15. Using this global
measure the .20 correlation of Factor E with the CERS scores and the .21
correlation of Factor C with the CERS scores seem worthy of more examina
tion. Likewise the correlations of Factors H and J with the BLRI scores

71
TABLE 9
MEAN CORRELATION OF
WITH ALL OF THE
TWO CRITERION
EACH HRPI FACTOR
SCORES OF THE
INSTRUMENTS
HRPI Factor
Mean of Correlation with All Scores
BLR I CERS
Factor
A
(Empathic Understanding -
Probing)
.16
.07
Factor
B
(Confrontation -
False Reassurance)
.01
.17
Factor
C
(Specificity Ambiguity)
.01
.21
Factor
D
(Analytical Interest -
Validation)
.05
-. 13
Factor
E
(Clarification -
Superficiality)
.06
.26
Factor
F
(Focused Probing -
Premature Judgments)
-.16
.04
Factor
G
(Tentative Formulations -
Avoidance of Emotions)
-.01
.00
Factor
H
(Requesting Giving
1nformation)
-.18
-.04
Factor
1
(Internal External Frame
of Reference)
-.03
.08
Factor
J
(Frank Feedback -
1ncongruence)
.18
.12
on the order of -.18 and .18 respectively merit consideration despite
the lack of precision afforded by this crude statistical treatment.

72
Testing the Hypotheses
Test of the First Major Hypothesis
The forward stepwise multiple regression technique was applied to
all scores of the two criterion instruments, the BLRI and the CERS,
using the ten factor scores of the HRPI as predictor variables. Four
demographic variables (age, grade point average, number of graduate
quarters completed, and number of quarters of supervised counseling
practice completed) were also entered into the computations as possible
predictors.
The first hypothesis postulated that higher client ratings could
be predicted from counselor trainee response preferences as measured
by the HRPI, the predictor instrument. Table 10 presents the predictors
that met the .05 level of significance required for inclusion in the
regression equation for each of the five BLRI scores. The strongest
predictor overall was graduate grade point average (GPA). In fact GPA
was the best predictor of the Total score in addition to three of the
subscores. GPA was the only variable which significantly predicted the
Congruence subscale. The GPA accounted for 8$ of the score variance in
the Level of Regard, the Empathy Level, and the Congruence scores. The
variance accounted for by the GPA was II# of the Unconditionality of
Regard score and rose to 13# of the Total score. No other demographic
variable was a significant predictor of any of the BLRI scores.
HRPI scores were significant predictors on four of the five BLRI
scores. Factor F (Focused Probing Premature Judgments) accounted for
5# of the variance in both the Total score and the Empathy Level score.
In both cases Factor F explained a lesser amount of variance than the

73
TABLE 10
BARRETT-LENNARD RELATIONSHIP INVENTORY PREDICTORS
Criterion Predictors
Multiple R
R2
df
F
Total Score
Graduate Grade Point Average
0.35
0.13
1,44
6.34*
Factor F (Focused Probing -
Premature Judgments)
0.42
0. 18
2,43
4.61*
Level of Reqard
Factor J (Frank Feedback -
1ncongruence)
0.41
0.17
1 ,44
8.70*
Graduate Grade Point Average
0.50
0.25
2,43
7.06*
Factor A (Empathic Under
standing Probing)
0.53
0.29
3,42
5.58*
Factor 1 (Internal External
Frame of Reference)
0.57
0.32
4,41
4.83*
Empathy Level
Graduate Grade Point Average
0.28
0.08
1 ,44
3.72*
Factor F (Focused Probing -
Premature Judgments)
0.36
0.13
2,43
3.28*
Unconditionality of Reqard
Graduate Grade Point Average
0.33
0. 1 1
1 ,44
5.44*
Factor J (Frank Feedback -
1ncongruence)
0.41
0. 17
2,43
4.42*
Conqruence
Graduate Grade Point Average
0.30
0.09
1,44
4.27*
*p<.05

74
primary predictor and only other predictor, the GPA. Also in each case
Factor F was a negative predictor.
The single strongest predictor in this study related to the Level
of Regard score. For that score Factor J (Frank Feedback Incongruence)
accounted for 17$ of the score variance. GPA added another 8$ to the
prediction process. Factor A (Empathic Understanding Probing) and
Factor I (Internal External Frame of Reference) each explained 4$
of the variance. As a result the multiple correlation rose to .57, the
highest in this research, and collectively the four predictors accounted
for 32$ of the Level of Regard score variance.
In addition to the 11$ of variance attributed to the GPA in the
Unconditionality of Regard score Factor J (Frank Feedback Incongruence)
again entered the equation in the amount of 6$ of variance accountability.
However, it should be noted that in this case, the variable which was
the strongest positive predictor of Level of Regard became a negative
predictor of Unconditionality of Regard.
Although the results are generally modest, ostensibly somewhat
contradictory, and involve only four of the ten HRPI factors, and only
four of the BLRI scores, there does appear to be a significant, obser
vable trend between client ratings and response preference scores on the
HRPI. Therefore, with due regard to the need for further clarifica
tion, the directional hypothesis HO| is accepted for the purposes of
this study.
Test of the Second Major Hypothesis
Even though the four demographic variables were also used in the
regression of the second criterion, supervisor ratings on the CERS, no

75
demographic variable, including GPA, showed up in the set of signifi
cant predictors found in Table II. It will be noted that for the CERS
Supervision score no predictor reached the necessary .05 level of
sign!ficance.
TABLE I I
COUNSELOR EVALUATION RATING SCALE PREDICTORS
(p<.05)
Criterion/Predictors
Multiple R
R2
df
F
Total Score
Factor E (Clarification -
Superficiality)
0.2721
0.0740
1 ,44
3. 52*
Counseling Score
Factor E (Clarification -
Superficiality)
0.3208
0.1029
1 ,44
5.05*
Factor D (Analytical Interest -
Va 1idation)
0.3836
0.1472
2,43
3.71*
Supervision Score
No Predictors
*p<.05
demographic variable, including GPA, showed up in the set of significant
predictors found in Table II. It will be noted that for the CERS Super
vision score no predictor reached the necessary .05 level of signifi
cance.
The CERS Total score was predicted by only Factor E (Clarification -

76
Superficiality) which accounted for 7.4$ of the variance. With only
one predictor the simple correlation and the multiple correlation were
an identical .27.
There were two factors predicting into the CERS Counseling score
but again the stronger predictor was Factor E with 10.3$ of the variance
and a simple correlation of .32. Factor D (Analytical Interest Vali
dation) added 4.4$ to the variance accounted for and raised the multiple
correlation to .38 with a total of 14.7$ of explained score variance.
The second major hypothesis that supervisor ratings on the Coun
selor Evaluation Rating Scale are predictable from HRPI response
preference scores, appears to be supported as far as Total score and
Counseling score are concerned. In this case only two HRPI subscores
are predicting the supervisor ratings, but attainment of the requi
site .05 confidence level of prediction allows the conclusion that
Factors E and D did successfully predict the criterion, supporting the
second major hypothesis with the limitations which have been noted.
It is possible, therefore, to construct regression equations to
predict the seven scores of the two criterion instruments. Table 12
presents the regression equations for predicting BLRI scores and Table
13 shows the regression equations for the two CERS scores predictable
from the variables under study here. The two regression equation tables
culminate this examination of the predictive potential of the Helper
Response Preference Inventory vis-a-vis client and supervisory ratings.

77
TABLE 12
REGRESSION EQUATION FOR THE BARRETT-LENNARD
RELATIONSHIP INVENTORY PREDICTORS
Dimensions
Equations
Tota 1 Score
80.
Level of Regard
=
2.
0.
44.
Empathy Level
=
19,
Unconditionality
of Regard
' 25.
Congruence
=
17.
10 (GPA) 3.62 (Factor F) 154.76
48 (Factor J) + 13.69 (GPA) +
53 (Factor A) 0.98 (Factor I) -
75
08 (GPA) 1.12 (Factor F) 32.44
13 (GPA) 2.01 (Factor J) 56.58
35 (GPA) 31.91

78
TABLE 13
REGRESSION EQUATION FOR THE COUNSELOR
EVALUATION RATING SCALE PREDICTORS
Scales Equations
Total Score

2.32 (Factor
E)
+ 140.10
Counseling Score
=
1,39 (Factor
E)
- 0.84 (Factor D)
+ 65.94
Supervision Score
(No Predictors)

CHAPTER V
DISCUSSION AND SUMMARY
The review of research on counseling effectiveness in Chapter II
demonstrated the varied and often circuitous routes to building a con
sistent model for selecting and training effective counselors. Never
theless, the centrality and interrelatedness of the counselors person
ality and communication style emerge consistently across research as fun
damental to the helping process. The literature review also suggested
that counselors may have characteristic manners of responding verbally
to clients. The propensity toward preferred responses is probably rooted
in both habit and value structure and may crucially shape the forma
tion of the therapeutic relationship as well as the outcome of the
helping process.
This study was designed to investigate the relationship of response
choices on the Helper Response Preference Inventory by counselor
trainees to client perceptions of the relationship and supervisor
evaluations. Client perceptions were assessed through the Barrett-
Lennard Relationship Inventory (BLRI) and supervisor ratings were
collected by means of the Counselor Evaluation Rating Scale (CERS).
It was hypothesized, first of all, that a relationship would be
found between counselor trainee response preferences on the HRPI and
client evaluations on the BLRI. Second, it was hypothesized that a
relationship would be found between response preferences on the HRPI
79

80
and supervisor evaluations on the CERS. This study also investigated
the degree of relationship among the three complete sets of variables
produced by the two criterion instruments and the predictor instrument.
A group of 46 graduate-1 eve I counselor trainees participated in
all aspects of the study along with their clients and supervisors. The
trainees completed the HRPI and were in turn evaluated by their clients
on the BLR I and by their supervisors on the CERS. By means of a step
wise regression technique the ten factor scores of the HRPI and four
selected demographic variables were regressed upon each of the criterion
instruments, employing both subscores and total scores. The findings
of this investigation were presented in Chapter IV.
Summary of the Results
Two directional hypotheses were tested during the investigation,
and each major hypothesis subsumed several hypotheses. The partial
questions break down each hypothesis into its components which are sub
scale and total scale scores within the criterion instrument. The major
hypotheses, the attendant partial hypotheses, and the results of the
analytical process on each are:
HO| There is a relationship between counselor trainee response
preferences and client evaluations of their counseling effec
tiveness. Fail to reject; partially.
1. A relationship exists between response preferences and Total
score on the BLRI. Fail to reject.
2. A relationship exists between response preferences and Level
of Regard score on the BLRI. Fail to reject.

81
3. A relationship exists between response preferences and Empathy
Level score on the BLRI. Fail to reject.
4. A relationship exists between response preferences and Uncon
ditionality of Regard score on the BLRI. Fail to reject.
5. A relationship exists between response preferences and Con
gruence score on the BLRI. Rejected.
HC>2 There is a relationship between counselor trainee response
preferences and supervisor evaluations of their counseling
effectiveness. Fail to reject; partially.
1. A relationship exists between response preferences and Total
score on the CERS. Fail to reject.
2. A relationship exists between response preferences and
Counseling score on the CERS. Fail to reject.
3. A relationship exists between response preferences and Super
vision score on the CERS. Rejected.
Practical Limitations
The nature of this research necessitates some consideration of the
actual limitations within which these findings must be interpreted.
In Chapter III it was pointed out that a deliberate sample limits
genera IizabiIity, that the external validity is compromised (somewhat)
by the exclusion of certain subjects, and that the primitive level of
development of the instruments involved makes dramatic assertions more
hazardous. In addition three more factors, in retrospect, may deserve
consideration as having influenced outcome.
An element of volunteerism was introduced when certain counselor
trainees chose not to participate in the research. The voluntary status

82
of involvement was accentuated by the failure of other counselor trainees
to follow through with participation to which they had earlier con
sented. Volunteerism was further introduced when 15 clients failed to
return their rating scale for trainees whose data sets were otherwise
complete. Such non-participation, regardless of the cause, tends to
limit the genera I izabi Iity of this investigation.
An artifact of the predictor instrument, the HRPI, is the restric
tion of choice range. The forced-choice format precludes any measure
ment of the creative capacity of the trainee to respond in a helpful
way. Theoretically, there could be a myriad of possible responses which
could be assigned to a wide spectrum of potential helpfulness. The
ability to respond creatively may not be measured, and conceivably
negated, by the format of the predictor instrument.
A third possible practical limitation is simply a part of the real
world, and therefore a potential hazard for any research utilizing
rating scales. Examination of the data suggests that raters, both
clients and supervisors, may have been influenced by certain ingrained
response patterns themselves. For example, the CERS employs a six-
point scale with three positive and three negative responses. One might
speculate that the most compatible response would be the middle positive
one which implies exceeding the minimum requirements but leaving room
for improvement. If such a phenomenon operated for all raters, the
Counseling and Supervision mean scores on the CERS would have been 78
and the Total score 162. The observed scores were 81.33, 84.22, and
171.54 respectively. Apparently a rating set did not occur in some
cases. Nevertheless, a fine-grained sensitivity to the actual nuances
of difference among counselors may have been influenced by this very

83
human pattern. The presence of significant observed differences may be
more remarkable when considered in this light.
Interpretation and Discussion of Findings
Although the two major hypotheses could not be accepted in full,
there were certain discernible trends apparent in the findings. Clearly
the HRPI factors are predicting into two different spheres represented
by client perceptions of the relationship on one hand and supervisor
evaluations of the counselors ability on the other. Six HRPI factors
achieved the .05 confidence level of significance as predictors. Four
predicted the BLRI scores and two predicted the CERS scores, but no
single factor predicted into both instruments.
The client perceptions registered on the BLRI appeared to place con
siderable value upon self disclosure. Frank Feedback was the single most
potent predictor in this study for any individual criterion score, pre
dicting the Level of Regard score. The same score also was predicted
in a lesser amount by responses from an external frame of reference which
in a sense supports the self disclosure theme. Along this same line the
BLRI Total score was predicted in part by premature judgments or quick
interpretations. The implication seems to be that clients perceive
honest, direct, and even potentially threatening feedback as concerned
involvement. In fact, in this study they experienced premature judg
ments, as opposed to focused probing, as a higher level of empathy.
It is worth noting that the frank honesty valued by clients did
not appear to be unbridled bluntness. For instance, empathic under
standing also was a significant predictor of Level of Regard. Despite
the potential threat of frank feedback from an external point of

84
reference, the communication style of the higher rated counselors
apparently was characterized by sensitivity and timing. The implica
tion is that the counselor cared enough to experience with and share
openly concerning the harsh, sometimes painful, realities of life.
Nevertheless, the helper transmitted, despite the uncomfortable con
tent, underlying caring for and commitment to the client.
It is interesting that Frank Feedback became a negative predictor
of the Unconditionality of Regard score. Perhaps this observation should
be interpreted in light of the growing disenchantment within counseling
theorist circles with the concept of unconditional regard. Some coun
selors question the possibility or even the appropriateness of communica
ting unconditional regard in the face of deteriorative or even destructive
client behavior. Certainly no definitive statement can be inferred from
this sparse evidence, but the field may continue to benefit from a
reexamination of the concept.
Unmistakably the most consistent predictor of BLRI scores was
graduate grade point average. This finding can be looked at in a number
of ways. In part the GPA had been influenced by practicum and intern
ship grades so the counselor trainee's GPA reflected some "expert"
evaluation of his or her counseling abiIity as we I I as a component of
cumulative experience. In part higher GPA suggests a higher verbal
facility which may be an indirect measure of the core concern of this
study, communication style. Another plausible explanation of the rela
tionship between high GPA and higher BLRI scores is a "prestige" factor.
Perhaps high GPA reflects an intelligence factor which impresses the
client. The counselor trainee with high GPA may be seen as one who is
knowledgeable, skilled, self-confident, and a worthwhile person with
whom to relate more closely.

85
A different pair of HRPI factors predicted two CERS scores. With
the supervisor criterion no demographic variable emerged as a predictor.
Also with the CERS, self disclosure did not appear to be meaningful
in the prediction process. The supervisor ratings were related to
clarification. Perhaps the supervisors rated positively the ability of
the counselor trainee to cut through superficial verbiage to highlight
implied meanings and lead the interchange into sharper focus. Clarifi
cation was the only predictor of the Total score and by far the best
predictor of the Counseling score. The supervisors evidenced that they
value the trainees skill in leading the client into deeper strata of
self understanding. This theme is supported by the only other predictor
of a CERS score. The Counseling score was also predicted by responses
that validate the client's experience. The implication is, "Yes, thats
the way you really are." The result is a very positive, affirming way
of developing the client's repertoire of personal insights.
Thus, in an overall perspective it can be said that the HRPI factor
scores predict both the BLRI and the CERS in divergent ways. For the
client data the response choices that favor frank self disclosure are
significantly predictive. For supervisor ratings it is those response
preferences that stress development of client self understanding which
predict at a meaningful level.
AppIication
The pertinent question that attends a study of this nature arose
in the introduction in Chapter I, persisted throughout Chapter II, and
in essence permeates this research. The question is whether the predic
tor instrument under investigation can or should be used for counselor

86
selection, training, or skill analysis. These questions will be con
sidered separately in light of the empirical findings.
The nature of a concurrent, cross-sectional investigation such as
this does not directly address the issue of counselor trainee selection.
To properly evaluate an instruments usefulness as a screening tool in
the candidate selection process a longitudinal study of the instru
ments predictive validity is necessitated. Nevertheless in the real
world of practical affairs those persons charged with the responsi
bility of candidate selection must base those decisions upon the best
available information. This instrument does provide some useful
information despite the modest variances accounted for, particularly
when used in conjunction with other indicators.
Perhaps the first question that must be resolved relates to which
definition of counseling effectiveness one wants to predict. Clearly
the HRPI predicts in different ways into both client and supervisor
definitions of effectiveness. Maybe the question is not an either-or
proposition since both criteria appear to be predictable from a single
test administration. Actually a candidate could be considered in the
light of his scores on the four factors which predict client ratings and
the two which predict supervisor ratings. One must do so, neverthe
less, with full awareness that thus far the instrument has not been
investigated in a manner thoroughly consistent with such use.
Whether the HRPI should be used for training or skill analysis is
another matter. While the predictive power is much too low to make life
changing decisions on this basis, a counselor trainee may gain some
insight into his own communication style through the instrument. For
example, if the counselor trainee's profile reveals a preference away

87
from frank feedback or clarification, he may be alerted to ways he can
enhance his effectiveness in terms of his clients perception, his super
visors perception, or both. Likewise, he may discover that his candor
is not balanced by understanding or vice versa, sensitizing himself to
areas worthy of emphasis in his ski I I-bui Iding process. In short, the
Helper Response Preference Inventory shows promise, albeit tentative,
as a useful educational tool to help bridge the gap between self dis
closure and empathic understanding and between the effectiveness models
of clients and supervisors.
Suggestions for Further Research
An examination of the findings of this study suggests that further
research in a number of areas may be beneficial. In general the ration
ale that gave rise to this study and the literature review in Chapter II
speak clearly of the need for development of instruments designed to
measure counseling effectiveness. Specifically there is a need for
instruments to assess the communication style of relatively effective
counselors in a continuing effort to tease out the elements distinctively
characteristic of helpful communication.
Previous references have been made to the need for a longitudinal
study to evaluate the effectiveness over time of counselor trainees who
have taken the h'RPI. The goal would be to arrive at a predictive formula
whereby a probability statement could be made regarding a counselor
trainee who attains a given score on the instrument. Of course, such
precision is a long way off and will doubtless prove to be elusive if
not unattainable. At the very least a thorough examination of how a
trainees HRPI score profile changes over time, especially with
practical supervised training, could be highly instructive.

88
Research into the relationship of the HRPI to other instruments
used to evaluate counseling effectiveness as outlined in Chapter II may
be profitable. Specifically it may be productive to relate response
preferences to the Rokeach Dogmatism Scale, for example, or to any of
the personality measures such as the 16 Personality Factor Questionnaire
or the Edwards Personal Preference Schedule. It is conceivable that a
multiple criterion approach to counseling effectiveness might yield a
combination of personality variables plus communicationaI modes which
would produce a predictive formula robust enough to survive replication.
Finally, this present research study should be replicated. Repli
cations should be carried out, first of all, under approximately equiva
lent conditions to examine the reiabiIity of these findings. Next it
would be useful to conduct similar investigations with differing samples
such as paraprofessionaIs and career helper trainees from various dis
ciplines, representing a broad spectrum of theoretical orientations.
Perhaps it would be enlightening to carry out this research design using
the same predictor instrument but substituting different client and
supervisor ratings as criterion instruments. Clearly, this study and
parallel studies have hardly scratched the surface of what may prove
to be a fertile field of investigation.
Conclusions of the Investigation
With due regard for the possible limitations of the study cited
in Chapter III and expanded in this present chapter, there are certain
tentative conclusions which can be drawn. First, the Helper Response
Preference Inventory does appear to predict client perceptions of the
counseling relationship on the BLRI in the case of four factors which

89
explain relatively modest portions of the score variance. Secondly,
the HRPI appears to predict supervisor ratings of the counselor trainee
on the CERS with two factors at a slightly more modest level. Thirdly,
the BLRI and the CERS appear to be measuring different aspects of the
counseling arrangement, and consequently the HRPI predicts into each
criterion in different ways with different response factors. Subse
quent use of the HRPI should be made in the light of these findings,
and future investigations in turn may amplify or modify the signifi
cance of these results.

APPENDICES

APPENDIX A
COUNSELOR TRAINEE DEMOGRAPHIC DATA SHEET
Your name:
Phone number:
Sex:
Age:
Undergraduate major:
Cumulative graduate grade point average at the end of last quarter:
Number of graduate quarters you will have completed at the end of this
quarter:
Number of quarters of supervised counseling practice (practicum or
internship) you will have completed at the end of this quarter:
Individual supervisor:
Present practicum or internship agency:
91

APPENDIX B
COVER LETTER TO THE CLIENT
To the cIient:
Your counselor has been asked to find one client who will rate him/her
as a counselor. You are the only client selected by your counselor to
take part in this research so your participation is extremely important.
You should know that how you rate your counselor will not be used to
grade or evaluate him/her personally. Rather all of the ratings will
be combined and examined as a group. Nevertheless, it is very impor
tant that we receive one hundred per cent cooperation. The name of
your counselor appears on your rating sheet only for the purpose of
identifying which rating scales have not been returned.
You will notice that the rating scale is stamped and addressed to make
your task as easy as possible. I realize it will take some time to
complete the Inventory, but your efforts will be sincerely appreciated.
Please read the instructions carefully before you begin.
Thank you,
Sidney L. Bradley
Doctoral Candidate
Department of Counselor Education
92

APPENDIX C
RELATIONSHIP INVENTORY-FORM OS-M-64
Below are listed a variety of ways that one person may feel or behave
in relation to another person.
Please consider each statement with reference to your present relation
ship with your
Mark each statement in the left margin, according to how strongly you
feel that it is true, or not true, in this relationship. Please mark
every one. Write in +3, +2, +1, or -I, -2, -3, to stand for the
following answers:
+3:
Yes, 1 strongly feel that it
-1 : No,
1 feel that it is probably
is true.
untrue or more untrue than true.
+2:
Yes, 1 feel it is true.
-2: No,
1 feel it is not true.
+ 1 :
Yes, 1 feel that it is
-3: No,
1 strongly feel that it is
probably true or more true
than untrue.
not
true.
1. He respects me as a person.
2. He wants to understand how I see things.
3. His interest in me depends on the things I say or do.
4. He is comfortable and at ease in our relationship.
5. He feels a true I iking for me.
6. He may understand my words but he does not see the way I feel.
7. Whether I am feeling happy or unhappy with myself makes no
real difference to the way he feels about me.
8. I feel that he puts oh a role or front with me.
9. He is impatient with me.
10. He nearly always knows exactly what I mean.
11. Depending on my behavior, he has a better opinion of me some
times than he has at other times.
12. I feel that he is real and genuine with me.
13. 1 feel appreciated by him.
14. He looks at what I do from his own point of view.
15. His feeling toward me doesnt depend on how I feel toward him.
16. It makes him uneasy when I ask or talk about certain things.
93

94
J7. He is indifferent to me.
18. He usually senses or realizes what I am feeling.
19. He wants me to be a particular kind of person.
20. I nearly always feel that what he says expresses exactly what
he is feeling and thinking as he says it,
_2I. He finds me rather dull and uninteresting.
22. His own attitudes toward some of the things I do or say
prevent him from understanding me.
23. I can (or could) be openly critical or appreciative of him
without really making him feel any differently about me.
24. He wants me to think that he likes me or understands me more
than he really does.
25. He cares for me.
26. Sometimes he thinks that J_ feel a certain way, because that's
the way he_ feels.
27. He likes certain things about me, and there are other things
he does not like.
28. He does not avoid anything that is important for our relation
ship.
29. I feel that he disapproves of me.
30. He realizes what I mean even when I have difficulty in saying
it.
31. His attitude toward me stays the same: he is not pleased
with me sometimes and critical or disappointed at other times.
32. Sometimes he is not at all comfortable but we go on, outwardly
ignoring it.
33. He just tolerates me.
J54. He usually understands the whole of what I mean.
35. If I show that I am angry with him he becomes hurt or angry
with me, too.
36. He expresses his true impressions and feelings with me.
37. He is friendly and warm with me.
38. He just takes no notice of some things that I think or feel.
39. How much he likes or dislikes me is not altered by anything
that I tell him about myself.
40. At times I sense that he is not aware of what he is really
feeling with me.
41. I feel that he really values me.
42. He appreciates exactly how the things I experience feel to me.

95
43. He approves of some things I do, and plainly disapproves of
others.
44. He is w¡I Iing to express whatever is actually in his mind with
me, including any feelings about himself or about me.
45. He doesnt like me for myself,
46. At times he thinks that I feel a lot more strongly about a
particular thing than I really do,
47. Whether I am in good spirits or feeling upset does not make
him feel any more or less appreciative of me.
48. He is openly himself in our relationship.
49. I seem to irritate and bother him.
50. He does not realize how sensitive I am about some of the
things we discuss.
51. Whether the ideas and feelings I express are "good" or "bad"
seems to make no difference to his feeling toward me.
52. There are times when I feel that his outward response to me
is quite different from the way he feels underneath.
53. At times he feels contempt for me.
54. He understands me.
55. Sometimes I am more worthwhile in his eyes than I am at other
times.
56. I have not felt that he tries to hide anything from himself
that he feels with me.
57. He is truly interested in me.
58. His response to me is usually so fixed and automatic that I
dont really get through to him.
59. I don't think that anything I say or do really changes the
way he feels toward me.
60. What he says to me often gives a wrong impression of his
whole thought or feeling at the time.
61. He feels deep affection for me.
62. When I am hurt or upset he can recognize my feelings exactly,
without becoming upset himself.
63. What other people think of me does (or would, if he knew)
affect the way he feels toward me.
64. I believe that he has feelings he does not tell me about that
are causing difficulty in our relationship.

APPENDIX D
RELATIONSHIP INVENTORYFORM OS-F-64
Below are listed a variety of ways that one person may feel or behave in
relation to another person.
Please consider each statement with reference to your present relation-
ship
with your
Mark each statement in the left margin, according to how strongly you
feel that it is true, or not true, in this relationship. Please mark
every one. Write in +3, +2, +1, or -1. -2, -3. to stand for the follow-
ing answers:
+3:
Yes, 1 strongly feel that it
is true.
-1 :
No, 1 feel that it is probably
untrue or more untrue than true.
+2:
Yes, 1 feel it is true.
-2:
No, 1 feel it is not true.
+ 1 :
Yes, 1 feel that it is
probably true or more true
than untrue.
-3:
No, 1 strongly feel that it is
not true.
1. She respects me as a person.
2. She wants to understand how I see things.
3. Her interest in me depends on the things I say or do.
4. She is comfortable and at ease in our relationship.
5. She feels a true liking for me.
6. She may understand my words but she does not see the way I
feel.
7. Whether I am feeling happy or unhappy with myself makes no
real difference to the way she feels about me.
8. I feel that she puts on a role or front with me.
9. She is impatient with me.
10. She nearly always knows exactly what I mean.
11. Depending on my behavior, she has a better opinion of me
sometimes than she has at other times.
12. I feel that she is real and genuine with me.
13. I feel appreciated by her.
14. She looks at what I do from her own point of view.
15. Her feeling toward me doesn't depend on how I feel toward her.
96

97
.16.
.17.
18.
.19.
20.
21
22.
23.
24.
25.
26.
27.
28.
29.
30.
.31 .
32.
33.
34.
.35.
36.
37.
38.
39.
40.
41 .
It makes her uneasy when I ask or talk about certain things.
She is indifferent to me.
She usually senses or realizes what I am feeling.
She wants me to be a particular kind of person.
I nearly always feel that what she says expresses exactly what
she is feeling and thinking as she says it.
She finds me rather dull and uninteresting.
Her own attitudes toward some of the things I do or say pre
vent her from understanding me.
I can (or could) be openly critical or appreciative of her
without really making her feel any differently about me.
She wants me to think that she I ikes me or understands me
more than she really does.
She cares for me.
Sometimes she thinks that J_ feel a certain way, because thats
the way she fee Is.
She likes certain things about me, and there are other things
she does not like.
She does not avoid anything that is important for our rela-
tionship.
I feel that she disapproves of me.
She realizes what I mean even when I have difficulty in saying
it.
Her attitude toward me stays the same: she is not pleased
with me sometimes and critical or disappointed at other times.
Sometimes she is not at all comfortable but we go on, out
wardly ignoring it.
She just tolerates me.
She usually understands the whole of what I mean.
If I show that I am angry with her she becomes hurt or angry
with me, too.
She expresses her true impressions and feelings with me.
She is friendly and warm with me.
She just takes no notice of some things that I think or feel.
How much she likes or dislikes me is not altered by anything
that I tell her about myself.
At times I sense that she is not aware of what she is really
feeling with me.
I feel that she really values me.

98
42. She appreciates exactly how the things I experience feel to
me.
43. She approves of some things I do, and plainly disapproves of
others.
44. She is willing to express whatever is actually in her mind
with me, including any feelings about herself or about me.
45. She doesnt like me for myself.
46. At times she thinks that I feel a lot more strongly about a
particular thing than I really do.
47. Whether I am in good spirits or feeling upset does not make
her feel any more or less appreciative of me.
48. She is openly herself in our relationship.
49. I seem to irritate and bother her.
50. She does not realize how sensitive I am about some of the
things we discuss.
51. Whether the ideas and feelings I express are "good" or "bad"
seems to make no difference to her feeling toward me.
52. There are times when I feel that her outward response to me
is quite different from the way she feels underneath.
53. At times she feels contempt for me.
54. She understands me.
55. Sometimes I am more worthwhile in her eyes than I am at other
times.
56. I have not felt she tries to hide anything from herself that
she feels with me.
57. She is truly interested in me.
58. Her response to me is usually so fixed and automatic that I
don't really get through to her.
59. I don't think that anything I say or do really changes the
way she feels toward me.
60. What she says to me often gives a wrong impression of her whole
thought or feeling at the time.
61. She feels deep affection for me.
62. When I am hurt or upset she can recognize my feelings exactly,
without becoming upset herself.
63. What other people think of me does (or would, if she knew)
affect the ways she feels toward me.
64. I believe that she has feelings she does not tell me about
that are causing difficulty in our relationship.

APPENDIX E
COUNSELOR EVALUATION RATING SCALE
Counselor Date
Supervisor
Below are listed some statements which are related to evaluation in
supervising a counseling experience. Please consider each statement
with reference to your knowledge of the counselor rated.
Mark each statement in the left hand blank according to how strongly
you agree or disagree. Please mark every statement. Write in +3, +2,
or +1, or -I, -2, or -3, to represent the following:
+3:
1 strongly agree
-1 :
1 s1ight1 y disagree
+2:
1 agree
-2:
1 disagree
+ 1 :
1 s1ight1y agree
-3:
1 strongly disagree
1. Demonstrates an interest in client's problems.
2. Tends to approach clients in a mechanical, perfunctory manner.
3. Lacks sensitivity to dynamics of self in supervisory relation-
shi p.
4. Seeks and considers professional opinion of supervisors and
other counselors when the need arises.
5. Tends to talk more than client during counseling.
6. Is sensitive to dynamics of self in counseling relationships.
7. Cannot accept constructive criticism.
8. Is genuinely relaxed and comfortable in the counseling session.
9. Is aware of both content and feeling in counseling sessions.
10. Keeps appointments on time and completes supervisory assign
ments.
11. Can deal with content and feeling during supervision.
12. Tends to be rigid in counseling behavior.
13. Lectures and moralizes in counseling.
14. Can critique counseling tapes and gain insights with minimum
help from supervisor.
15. Is genuinely relaxed and comfortable in the supervisory session.
99

100
16. Works well with other professional personnel (e.g. teachers,
counselors, etc.)
17. Can be spontaneous in counseling, yet behavior is relevant.
18. Lacks self-confidence in establishing counseling relation
ships.
19. Can explain what is involved in counseling and discuss
intelligently its objectives.
20. Is open to self-examination during supervision.
21. Can express thoughts and feelings clearly in counseling.
22. Verbal behavior in counseling is appropriately flexible and
varied, according to the situation.
23. Lacks basic knowledge of fundamental counseling principles
and methodology.
24. Participates actively and willingly in supervisory sessions.
25. Is indifferent to personal development and professional
growth.
26. Applies a consistent rationale of human behavior to counseling.
27. Can be recommended for a counseling position without reserva
tion.
Recommended Grade
Comments:

APPENDIX F
HELPER RESPONSE PREFERENCE INVENTORY
Pi reef ions: This instrument is designed to measure your preference for
various types of responses which one person (the helper) might reasonably
make to another (the helpee) in an effort to assist him to solve his
problems. The Inventory consists of typical helpee statements each
followed by a pair of possible helper responses. For each helpee state
ment, select the helper response which you believe would be most appro
priate.
Be sure to select only one response for each helper
statement.
Do not leave any item blank.
You should assume that each of the helpee statements
was made by a different person.
Mark your preference on the answer sheet provided.
Do not mark or write on the inventory.
I.Helpee: "I think I really should talk with someone about my
problems, but I'm not so sure that I can."
Helper A; "What kind of problems do you seem to be having?"
Helper B: "You really shouldn't be upset about discussing your
problems. Everyone has difficulties."
2.Helpee: "No matter how hard I study, I seem to keep making silly,
stupid mistakes in math!"
Helper A: "What kind of grades have you been getting in your math
course?"
Helper B: "This must be pretty frustrating for you after working
at it so."
3.Helpee: "My wife and I were getting along pretty well until her
first pregnancy, but things have been pretty much going
downhill since then."
Helper A: "It sounds like you may have resented her pregnancy."
Helper B: "That isn't an unusual reaction to the first pregnancy."
4.Helpee: "I really had expected to see a psychiatrist when I
made this appointment. You're a social worker, aren't
you?"
101

102
Helper A:
"I'm sure I'll be able to help you."
Helper B:
"You're pretty concerned about status, aren't you?"
5. Helpee:
"1 feel so good, lately, now that I'm beginning to see
how I've been bringing some of these things on myself."
Helper A:
"You seem to be enjoying 1 ife a little more, now that
you feel that you understand how some of your diffi
culties have been coming about."
Helper B:
"Life really begins to mean something when you face
your responsibilities, doesn't it?"
6. Helpee:
"1 don't see how you can manage to sif and listen to me
week after week without being bored out of your mind."
Helper A:
"But you really don't bore me at all."
Helper B:
"You seem to feel that there's nothing about you that
anyone could find interesting."
7. Helpee:
"My scores on all of the tests I've taken so far point
toward success in sales work, but every time 1 just
about get up enough courage to switch majors, 1 get cold
feet and back out."
Helper A:
"You seem to be drawn to selling becuase of the rewards
you might realize there, but some of the interpersonal
skills required are causing you to re-evaluate your
decisi on."
Helper B:
"Logically, you have some evidence that this is some
thing you could do, but somehow you feel that there's
something about this that isn't quite right for you."
8. Helpee:
"My friends get all upset just because I'm a pretty
heavy drinker. 1 don't see how that should concern
them, just so long as 1 don't get in their way."
Helper A:
"It's really up to you, isn't it? If it bothers them,
that's really their problem, so long as you can handle
it."
He 1 per B:
"But you can't live your life alone. You ought to think
of how you're affecting other people through your
behavior."
9. Helpee:
"1 just feel so alone right now, just like I'm back
inside a shell of some kind with no way to reach out."
Helper A:
"But you really aren't alone, and you can reach out.
People will respond."

103
Helper B:
"What kinds of things seem to lead up to the feelings
youre describing now?"
10. Helpee:
"Ive been a teacher for seventeen years, and Ive
always put the kids first, and look where that's got
me!"
Helper A:
"But you really have been right, you know. You really
should put the growth of the children first."
Helper B:
"You seem to be rationalizing your current situation by
saying that you haven't been appreciated for being a
good teacher."
II. He 1 pee:
"My parents are pretty determined that 1 should stay at
home and attend the local junior college. 1 feel that
I'm being headstrong and rebellious, but 1 just cant
live at home any longer, college or no college!"
Helper A:
"It's probably best for you to get out, since you feel
so strongly about it."
Helper B:
"What seems to be the trouble at home? Are your parents
not getting along with each other?"
12. Helpee:
"If the boss leans on me one more time like he did
today, I'm going to give it to him straight. I've just
about had it!"
Helper A:
"1 know how you feel, but you'll probably feel a lot
better after you've blown off a little of the steam to
me."
Helper B:
"You feel pretty sore at him, don't you?"
13. Helpee:
"1 don't see why 1 can't do it. All of my friends are."
Helper A:
"But, 1 don't want you to feel bad about it."
Helper B:
"You seem to want to justify your behavior through the
actions of others."
14. Helpee:
"No matter how hard 1 try, 1 still don't feel that 1
measure up or that people feel that I'm good enough for
them."
Helper A:
"But others may not feel that way about you."
Helper B:
"It's not good for you to let others influence you so
much."
15. Helpee:
"Why should 1 study harder? No matter how well prepared
1 am, 1 still don't have a fighting chance against the
frat guys who always have copies of the exams."

104
Helper A:
"How do you know that this cheating is going on?"
Helper B:
"You're probably afraid to put forth real effort for
fear of failure."
16. Helpee:
"Im really surprised at these test results! 1 never
expected to have such a high scientific interest score."
Helper A:
"Why do these results surprise you so much?"
Helper B:
"Yes, your score is quite high, very consistent with
your good grades in physics and chemistry."
17. Helpee:
"I've never felt comfortable when I'm with people. Now
that I'm having to do more public speaking, it's really
getting to me."
Helper A:
"It's getting pretty frustrating, isn't it?"
Helper B:
"It's probably worse now that you're in a position where
people may be evaluating you."
18. Helpee:
"I've thought about the pros and cons, and 1 still don't
know which way 1 should go. What do you think 1 should
do?"
Helper A:
"You should be depending on your own judgment, not mine."
Helper B:
"You seem uneasy about facing the consequences of either
course of action."
19. Helpee:
"My husband is a fine person and good to the children,
so 1 guess it isn't fair for me to criticize him."
Helper A:
"You dislike some things about him, but feel guilty
about it."
Helper B:
"What do you think would happen if you were to express
some of your feelings to him?
20. Helpee:
"I've thought a lot about how 1 affect other people. 1
know that I've had a lot of trouble, but it's hard to
see just how 1 bring it on myself, if that's what really
happens."
Helper A:
"But you should examine your own behavior closely."
Helper B:
"You feel that you really aren't at fault much of the
time?"
21. Helpee:
"1 won't sit down! All we ever do is sit and listen
while you talk to us!"

105
Helper A:
"Youre angry because 1 had to punish you yesterday!"
Helper B:
"Please sit down. Were going to have a nice surprise!"
22. He 1 pee:
"Why does he do that to me? 1 can't understand how he
can keep it up?"
Helper A:
"You really resent it, don't you?"
Helper B:
"Maybe he really doesn't intend to hurt you."
23. Helpee:
"I've never tried to smoke pot, but most of my friends
do. 1 realize it's just a matter of time before 1
check it out."
Helper A:
"Of course, you know the risks you would be taking if
you give in to their pressures."
Helper B:
"What do you expect to get out of it, when you try it?"
24. Helpee:
"Mrs. Brown, I've finished my math assignment, but Sally
thinks 1 haven't done it right. Can 1 look it over
again and turn it in tomorrow?"
Helper A:
"No, Mary. 1 want you to give it to me now. You know
it's important for you to do your own work."
Helper B:
"I'd like you to turn it in now, Mary. Let me look it
over and try not to worry about the mistakes."
25. Helpee:
"1 think 1 want to sign up again for dramatics. You
know, 'cause that's the one thing in school so far that
really turns me on."
He 1 per A:
"This is one thing you really like."
Helper B:
"Maybe you like dramatics because it gives you more of
a chance to express yourself."
26. Helpee:
"Last night 1 had this dream in which 1 was a drill
sergeant, and 1 was getting a big kick out of shoving
the recruits around."
Helper A:
"Maybe you have a greater need for power than youve
realized up to now."
Helper B:
"Can you think of some real life situations where you
get that same satisfaction from being in charge?"
27. Helpee:
"1 came from a very happy home. In fact, there was
never a quarrel or disagreement between my parents, but
my wife and 1 seem to be fighting constantly."

106
Helper A:
"Perhaps there was more conflict in your family than you
realized at the time."
Helper B:
"Its good that youve recognized the trouble before it
got too serious."
28. Helpee:
"Every day 1 go to class, 1 get this panicky feeling,
just like I'm losing my mind or something like that."
Helper A:
"Tell me, if you can, where you get this feeling most
often?"
Helper B:
"This must be pretty frightening for you."
29. Helpee:
"After the depression, 1 never got back on my feet
financially, and it seems that 1 just never could make
ends meet after that."
Helper A:
"So things have been tough for quite a long time."
He 1 per B:
"Well, let's hope things will be better shortly."
30. Helpee:
"When 1 left the state university last year, 1 was real
confident that 1 could make up my mind about what 1
wanted to do. But the longer I'm out, the more confused
things seem."
Helper A:
"At the time you left, what specific plans had you
formu1ated?"
Helper B:
"It often takes some time to really decide what you do
want to do, especially right after you finish college."
31. Helpee:
"1 don't believe much in religion any more. I've
gradually come to realize that most people simply use
religion as a crutch to compensate for their own weak
nesses."
Helper A:
"But I'm sure you know that some strong persons really
do have deep religious convictions."
Helper B:
"Do you mean that religion hasn't really been of help
to you, that you've had to work things out for yourself?"
32. Helpee:
"1 did well on the college entrance exam and the reading
test, but my math aptitude scores were way below average.
Do you think that wi11 cause problems for me?"
Helper A:
"1 don't think so unless you plan a program requiring a
lot of math. This suggests that you've either had little
work in math or that you really didn't find it very
cha 1lenging."

107
Helper B:
"1 really doubt it. With your good scores in general
aptitude and reading, you should do well in almost
everything besides math."
33. Helpee:
"Its finally coming clear to me. 1 had always believed
that people were usually disappointed in me, but now it
doesn't seem that that's been the case."
Helper A:
"So, now you're having some different ideas about how
people react to you."
Helper B:
"It must be a good feeling to find that people haven't
been so disappointed in you all along."
34. Helpee:
"My principal doesn't seem to understand what I'm trying
to do in this course. I'm afraid he won't recommend my
appointment for next year."
Helper A:
"He shouldn't drop you just because he doesn't under
stand what you're doing."
Helper B:
"But this is early in the year. I'm sure he'll feel
differently long before the contracts are issued."
35. Helpee:
"My friends really don't understand me. We have a lot
of laughs together, but they don't really know what I'm
1ike."
Helper A:
"Gets kinda lonely at times?"
Helper B:
"What are the things about you that you think your
friends don't know?"
36. Helpee:
"I'm afraid to go home anymore. It seems like there's
always an argument with someone yelling or screaming at
you."
Helper A:
"This must be pretty unpleasant for you."
Helper B:
"Is this because people in your home aren't able to
really communicate with one another?"
37. Helpee:
"The psychologist who gave me a 11 of those tests would
n't tell me what the results were. Could you give me
some indication of how 1 did?"
Helper A:
"Were there some tests in particular that you were
worried about?"
Helper B:
"We may be able to discuss your results, but, you know,
these aren't the kinds of tests you pass or fail."

108
38. Helpee:
"My brother was a lot older than 1, and we never had a
lot in common, but 1 always looked up to him, a lot more
than 1 did to my father."
Helper A:
"How did you get along with your father?"
Helper B:
"Maybe your brother really acted as a father toward
you."
39. Helpee:
"1 want to take my comprehensive exams this term, but
1 may decide to wait. Some of the people writing this
time are the real top students, and 1 might have a
better chance to pass next time."
Helper A:
"But you've prepared well for the exams and have a good
chance of passing, probably better than you give your-
self credit for."
Helper B:
"1 think you may be avoiding really finding out what
you are able to do on the exams."
40. Helpee:
"We've talked several times now, but things don't seem
to be changing much. Do you think we're getting any
where?"
Helper A:
"But this is the way we have to continue in order for
you to make the kind of progress you want."
Helper B:
"Are you beginning to wonder if this is a waste of time?"
41. Helpee:
"When is mommie coming to get me?"
Helper A:
"I'm sure she'll be here very soon."
Helper B:
"It'll be good when she gets here, won't it?"
42. Helpee:
"1 haven't taken anything from a store now for almost
two months. I'm hoping this shoplifing binge is over
for good."
Helper A:
"Have you had the urge to pick up things the last few
times you've been shopping?"
Helper B:
"That's a real improvement. I'm glad for you."
43. Helpee:
"Mr. Smith, you don't know how surprised 1 was when 1
found out that I'd passed your course. 1 know my grades
were awful, and 1 was almost positive that Id goofed
up the final exam."
Helper A:
"Were you surprised because you thought you were failing
up to the point of the exam?"

109
Helper B:
"1 know its been hard for you, but 1 dont feel that
you've ever resorted to deliberate deception.
45. Helpee:
"Ive been engaged for six months now, and my fiance
wants to live together, even though we wont be able to
get married until we both graduate. I'm really in con
flict over what to do."
Helper A:
"Is this partly because you are afraid of what he'll
think of you if you dont give in?"
Helper B:
"It hasn't been an easy thing to decide, has it?"
46. Helpee:
"1 wish 1 could have more self confidence. Everyone 1
know seems to have it."
Helper A:
"What are the things you don't feel confident about?"
Helper B:
"You feel this is something missing that you really
want."
47. Helpee:
"Just because 1 failed to promote six kids last year,
that doesn't mean I'm not a good teacher. But 1 suppose
that's just not the way to get along around here."
Helper A:
"Are you confident that the kids you failed really
weren't making the grade?"
Helper B:
"Could you be using the failures as the reason for the
negative feedback you're getting?"
48. Helpee:
"Do you think I'll ever be able to pass this course?
I've just looked over the texts and the reading list,
and I'm afraid that it's just too much for me."
Helper A:
"1 suspect that most students have that same reaction at
the beginning of the course. But you realize that most
of them do actually pass the course, and some do very
we 11."
Helper B:
"1 wonder why you feel this way. Have you had some bad
experiences with this subject in high school?"

BIBLIOGRAPHY
Abeles, N. A study of the characteristics of counselor trainees. Pis-
sertation Abstracts, 1958, 18, 2204-2205.
Abramowitz, S. I., & Jackson, C. The comparative effectiveness of
there-and-then versus here-and-now therapist interpretations in
group psychotherapy. Journal of Counseling Psychology, 1974, 21,
288-293.
ACES Committee on Standards for Counselor Education Secondary.
Standard for the preparation of secondary school counselors 1967
edition. Memorandum to ACES members, January 0, 1967.
Allen, T. W. Effectiveness of counselor trainees as a function of
psychological openness. Journal of Counseling Psychology, 1967,
14, 35-40.
American Personnel and Guidance Association. A statement of policy.
The counselor: Professional preparation and role. Personnel and
Gu i dance Journa I 1963, 4J_, 480-485.
American Psychological Association, Education and Training Board.
Criteria for evaluating training programs in clinical and counsel
ing psychology. American Psychologist, 1958, _P3, 59-60.
American Psychological Association, Subcommittee on Counselor Trainee
Selection, Counselor Training Committee, Division of Counseling
Psychology. An analysis of practices in counselor trainee selec
tion. Journal of Counseling Psychology, 1954, 1, 174-179.
Anderson, R. P., & Anderson, G. V. Development of an instrument for
measuring rapport. Personnel and Guidance Journal, 1962, 41, 18-
24.
Arbuckle, D. S. Client perception of counselor personality. JournaI
of Counseling Psychology, 1956, _3, 93-96.
Arbuckle, D. S., & Wicas, E. A. The development of an instrument for
the measurement of counseling perceptions. Journal of Counseling
Psychology, 1957, 4_, 304-310.
Armstrong, J. C. Perceived intimate friendship as a quasi-therapeutic
agent. Journal of Counseling Psychology, 1969, J_6, 137-141.
I 10

Bailey, M. A., Warshaw, L., & Eichler, R. M. A study of factors related
to length of stay In psychotherapy. Journal of Clinical Psychology,
1959, J_5, 442-444.
Barahal, D. G., Brammer, L. M., & Shostrom, E. L. A client-centered
approach to educationaI-vocationaI counseling. Journal of Con
sulting Psychology, 1950, J_4, 256-260.
Barry, R., & Wolf, B. Five years of the Personnel and Guidance Journal.
Personnel and Guidance Journal, 1958, 36, 549-556.
Barrett-Lennard, G. T. Dimensions of therapist response as causal
factors in therapeutic change. Psychological Monographs, 1962,
76, No. 43.
Barrett-Lennard, G. T. Technical note on the 64-item revision of the
Relationship Inventory. Unpublished mimeograph report, Univer
sity of Waterloo, 1969.
Barrington, B. L. Prediction from counselor behavior of client percep
tion and of case outcome. Journal of Counseling Psychology, 1961,
8, 37-42.
Bartlett, W. E., & Thompson, C. L. Counselor preparation: A semantic
differential evaluation. Counselor Education and Supervision,
1971, JJ_, 129-133.
Bateson, G., Jackson, D. 0., Haley, J., & Weakland, J. Toward a
theory of schizophrenia. Behavioral Science, 1956, J_, 251-264.
Benton, J. A. Perceptual Characteristics of Episcopal Pastors.
Doctoral dissertation, University of Florida, 1964.
Berenson, B. G., & Carkhuff, R. R. The Sources of Gain in Counseling
and Psychotherapy. New York: Holt, Rinehart and Winston, Inc.,
1967.
Berenson, B. G., Carkhuff, R. R., & Myrus, P. The interpersonal func
tioning and training of college students. Journal of Counseling
Psychology, 1966, JJ5, 441-446.
Bergin, A. E., & Solomon, S. Personality and performance correlates
of graduate training. American Psychologist, 1963, _l_8, 393.
Berzon, B. The self-directed therapeutic group: An evaluative study.
LaJolla, California: Western Behavioral Sciences Institute
Reports, 1964, No. I
Bishop, J. Another look at counselor, client, and supervisor ratings
of counselor effectiveness. Counselor Education and Supervision,
1971, _[0, 319-322.

Blocher, D. H. A multiple regression approach to predicting success
in a counselor education program. Counselor Education and Super
vision, 1963, 3, 19-22.
Blumberg, A. Supervisory behavior and interpersonaI relations. Educa
tional Administration Quarterly, Spring, 1968, 34-45.
Board, F. A. Patients and physicians' judgments of outcome of psycho
therapy in an outpatient clinic. Archives of General Psychiatry,
1959, i, 185-196.
Bordin, E. S. Implications of client expectation for the counseling
process. Journal of Counseling Psychology, 1955, 2, 17-21.
Blocker, D. H. A multiple regression approach to predicting success in
a counselor education program. Counselor Education and Super
vision, 1963, 3, 19-22.
Bordin, E. S. Implications of client expectation for the counseling
process. Journal of Counseling Psychology, 1955, 2^, 17-21.
Bordin, E. S., Cutler, R. L., Dittmann, A. T., Harway, N. L., Raush,
H. L., & Rigler, D. Measurement problems in process research in
psychotherapy. Journal of Consulting Psychology, 1954, 18,
79-85.
Borman, C. A., & Ramirez, C. Evaluating counseling practicum students.
Counselor Education and Supervision, 1975, 15, 48-54.
Bowler, E. M., & Dawson, F. T. Counseling Employees. New York:
Prentice-Hall, 1948.
Bradley, F. 0. A modified interpersonal process recall technique as a
training model. Counselor Education and Supervision, 14, 34-39.
Bradley, S. L. Examination of the Helper Response Preference Inven
tory. Unpublished mimeograph report, University of Florida, 1973.
Bradley, S. L. Factor analysis of the Helper Response Preference
Inventory. Unpublished mimeograph report, University of Florida,
1973.
Brammer, L. M., & Shostrom, E. L. Therapeutic Psychology. Englewood
Cliffs, N. J.: Prentice-Hall, Inc., I960.
Brams, J. M. Counselor characteristics and effective communication in
counseling. Journal of Counseling Psychology, 1961, 25-30.
Brown, D., & Cannaday, M. Counselor, counselee, and supervisor ratings
of counselor effectiveness. Counselor Education and Supervision,
1969, 8, I 13-118.
Brown, 0. B., & Cala, V. F. Two methods of initiating student

Interviews: Self-initiated versus required. Journal of Counseling
Psychology, 1968, _[5, 402-406.
Brown, R. G. A. study of the perceptua I organ i zat ion of e I ementary and
secondary Outstanding Young Educators. Doctoral dissertation,
University of Florida, 1970.
Burnett, C. W. Selection and training of school and college personnel
workers. Review of Educational Research, 1954, 24, 121-133.
Burstein, J. W., & Carkhuff, R. R. Objective therapist and client
ratings of therapist-offered fcilitative conditions of moderate
to low functioning therapists. Journal of Clinical Psychology,
1968, 24, 240-241.
Cahoon, R. A. Some counseI or attitudes and characteristies re Iated to
the counseling relationship. (Doctoral dissertation, Ohio State
University) Ann Arbor, Michigan: University Microfilms, 1962.
No. 63-2480.
Cannon, J. R., & Carkhuff, R. R. Effects of rater level of function
ing and experience upon the discrimination of fcilitative condi
tions. Journal of Consulting and Clinical Psychology, 1968, 33,
189-194.
Carkhuff, R. R. Helping and human relations, volume I, selection and
training. New York: Holt, Rinehart, and Winston, Inc., 1969.
(a)
Carkhuff, R. R. Helping and human relations: A primer for lay and
professional helpers, voIume I I: Practice and research. New
York: Holt, Rinehart, and Winston, Inc., 1969. (b)
Carkhuff, R. R. Training in the counseling and therapeutic practices:
Requiem or rev¡elle. Journal of Counseling Psychology, 1966, 13,
360-367.
Carkhuff, R. R., & Alexik, M. Effect of client depth of self-explora
tion upon high- and low-functioning counselors. Journal of
Counseling Psychology, 1967, J_4, 350-355.
Carkhuff, R. R., & Berenson, B. G. Beyond counseling and therapy.
New York: Holt, Rinehart, and Winston, Inc., 1967.
Carkhuff, R. R., Collingwood, T., & Renz, L. The effects of didactic
training upon trainee level of discrimination and communication.
Journal of Clinical Psychology, 1969, 25, 460-461.
Carkhuff, R. R., Kratochvil, D., & Friel, T. Effects of professional
training: Communication and discrimination of fcilitative condi
tions. Journal of Counseling Psychology, 1968, 15, 68-74.

Carkhuff, R. R., Piaget, G., & Pierce, R. The development of skills
in interpersonal functioning. Counselor Education and Supervision,
1968, 7, 102-106.
Carkhuff, R. R., & Truax, C. B. Training in counseling and psycho
therapy: An evaluation of an integrated didactic and experiential
approach. Journal of Consulting Psychology, 1965, 29, 333-336.
Carkhuff, R. R., & Truax, C. B. Toward explaining success and failure
in interpersonaI learning experiences. Personnel and Guidance
Journal, 1966, 44, 723-727.
Carson, R. C., Harden, J. A., & Shows, W. D. A-B distinction and
behavior in quasi-therapeutic situations. Journal of Consulting
Psychology, 1964, 28, 426-433.
Cartwright, R. D., & Lerner, B. Empathy, need to change, and improve
ment with psychotherapy. Journal of Consulting Psychology, 1963,
27, 138-144.
Cartwright, R. D., & Vogel, J. L. A comparison of changes in psycho
neurotic patients during matched periods of therapy and no therapy.
Journal of Consulting Psychology, I960, 24, 121-127.
Cash, W. L., Jr., & Munger, P. F. Counselors and their preparation.
Review of EducationaI Research, I 966, 36, 256-263.
Cattell, J. M. Statistics of American psychologists. American Journal
of Psychologists, 1903, J_4, 310-328.
Chap line, E. B. A study of cognitive complexity, selected demographic
variabIes and counseIinq behavior of counselor-trainees. (Doc
toral dissertation, Temple University) Ann Arbor, Mich.: Univer
sity Microfilms, 1964. No. 64-9329.
Chinsky, J. M., & Rappaport, J. Brief critique of the meaning and
reliability of "accurate empathy" ratings. Psycho logical Bulletin,
1970, 73, 379-382.
Colby, K. M. Computer simulation of a neurotic process. In Tomkins,
S. S. & Messick, S. (Eds.) Computer Simulation of Personality.
New York: John Wiley, 1963.
Combs, A. W. A perceptual view of "helpers" in personality theory and
counseling practice. First Annual Conference on Personality
Theory and Counseling Practice. Gainesville: University of
Florida Press, 1961.
Combs, A. W., Avila, D. W., & Purkey, W. W. Helping relationships:
Basic concepts for the helping professions. Boston: AI Iyn-Bacon,
1971. '
Combs, A. W., & Snygg, D. Individual behavior. New York: Harper and
Row, 1959.

Combs, A. W., & Soper, D. W. The perceptual organization of effective
counselors. Journal of Counseling Psychology, 1963, 10, 222-227.
Cornell, P. T. Factors influencing communication in counseling.
Doctoral dissertation, University of Missouri, 1955.
Cottle, W. C. Personal characteristics of counselors. Personnel and
Guidance JournaI, 1953, 31, 445-449.
Cottle, W. C., Lewis, W. W., Jr., & Penny, M. M. Personal characteris
tics of counselors. I I I An experimental scale. Journal of Coun
seling Psychology, 1954, J_, 74-77.
Cottle, W. C., & Lewis, W. W., Jr. Personality characteristics of
counselors: II Male counselor responses to MMPI and GZTS.
Journal of Counseling Psychology, 1954, J_, 27-30.
Cox, Rachel. Counselors and their work. Philadelphia: Archives
Publishing Company, 1945.
Cronbach, L. J. Essentials of psychological testing. New York:
Harper and Row, I960.
Davis, R. M., & OConner, J. B. Changes in counseling response prefer
ence among rehabilitation counseling interns. Counselor Educa
tion and Supervision, 1974, J_4, 40-46.
Dedrick, C. The relationship between perceptual characteristics and
effective teaching at the junior college level. Doctoral disser
tation, University of Florida, 1972.
Demos, G. D., & Zuwaylif, F. H. Characteristics of effective counselors.
Counselor Education and Supervision, 1966, 5_, 163-165.
Dickman, J. The perceptual organization of person oriented versus task
oriented student nurses. Doctoral dissertation, University of
Florida, 1967.
Dilley, J. Supervisory ratings of counselor trainees in a simulated
work setting as compared with peer and instructor ratings of the
same trainees in an academic setting. Counselor Education and
Supervision, 1964, ~b_, 70-73.
Dole, A. A. The prediction of effectiveness in school counseling.
Journal of Counseling Psychology, 1964, JJ_, I 12-121.
Donnan, H. H., Harlan, G. E., & Thompson, S. A. Counselor personality
and level of functioning as perceived by counselees. Journal of
Counseling Psychology, 1969, J_6, 482-485.
Doy le, E. J. The relationship between coI Iege teaching effectiveness
and inferred characteristics of the adeguate personality. Doctora I
dissertation, University of Northern Colorado, 1969.

116
Driekurs, R., & Sonstegard, M. A specific approach to practicum super
vision. Counselor Education and Supervision, 1966, 6^ 18-26.
Dunlop, R. S. Counselor competence: Some proposals in search of
advocacy. Personnel and Guidance Journal, 1968, _46, 655-660.
Ebel, R. L. Estimation of the reliability of ratings. Psychometrika,
1951, 16, 407-424.
Eberlein, L., & Park, J. Se If-concept/idea I-self-concept congruence
and rated effectiveness of counselor trainees. Counselor Educa
tion and Supervi sion, 1971, J_0, 26- 132.
Embree, R. B. The assessment of counselor competence through the use
of multiple ratings by supervisors and fellow students. American
Psycho I oqi st, 1954, 9_, 360.
Emmerling, F. C. A study of the relationship between personality
characteristies of classroom teachers and pup i I perceptions of
these teachers. (Doctoral dissertation, Auburn University) Ann
Arbor, Michigan: University Microfilms, 1961.
FauIkenberry, P. D. A study of the evaluation of counseling effective
ness based upon supervision, peer, role pIayer, and self-ratinqs.
(Doctoral dissertation, Purdue University) Ann Arbor, Mich.:
University Microfilms, 1968. No. 68-12, 550.
Fiedler, F. E. Factor analysis of psychoanalytic, nondirective, and
adlerian therapeutic relationships. Journal of Consulting
Psychology, 1951, J_5, 32-38.
Fiedler, F. E. The concept of an ideal therapeutic relationship.
Journal of Consulting Psychology, 1950, J_4, 239-245.
Flanagan, J. C. The critical incident techniques. Psychological
Bulletin. 1954, 51, 327-358.
Foley, W. J., & Proff, E. C. NDEA institute trainees and vocational
rehabilitation counselor: A comparison of characteristics.
Counselor Education and Supervision, 1965, 1, 154-159.
Ford, D. H., & Urban, H. B. Psychotherapy. In Farnsworth, P. R.,
(Ed.) Annual Review of Psychology, Vol. 13, Palo Alto, California:
Annual Reviews, 1967.
Forgy, W. E., & Black, J. D. A follow-up after three years of clients
counseled by two methods. Journal of Counseling Psychology, 1954,
1, 1-8.
Form, A. L. The construction of a scale on attitudes toward counseling.
Journal of Counseling Psychology, 1955, 7_, 96-102.
Forster, J. R., & Flamburg, R. L. Further exploration of the 16 PF

and counselor effectiveness. Counselor Education and Supervision,
1976, J_5, 184-188.
Freedman, S. A., Antenen, W. W., & Lister, J. L. Counselor behavior
and personality characteristics. Counselor Education and Super-
vi sion, 1967, 26-30.
Friesen, D. D., & Dunning, G. B. Peer evaluation and practicum super
vision. Counselor Education and Supervision, 1973, 12, 229-235.
Froehlich, C. P. Preparation of teachers and specialists for guidance
service. Review of Educational Research, 1951, 21, 159-166.
Gade, E. M. The relationship of sociometric indices and counselor
candidate effectiveness. Counselor Education and Supervision, 1967,
6, 121-124.
Garrison, M. A correlation study of factors utilized in counselor
selection. Unpublished specialist project, Western Michigan
University, 1967.
Gooding, C. T. An observational analysis of the perceptual organiza
tion of effective teachers. Doctoral dissertation, University of
Florida, 1974.
Goodstein, L. D., & Grigg, A. F. Client satisfaction, counselors and
the counseling process. Personnel and Guidance Journal, 1959, 38,
19-24.
Graves, J. T. The employee counselor. Occupations, 1944, 22, 495-497.
Grigg, A. E. Client response to counselors at different levels of
experience. Journal of Counseling Psychology, 1961, 8^ 217-223.
Grigg, A. E., & Goodstein, D. D. The use of clients as judges of the
counselor's performance. Journal of Counselinq Psychology. 1957,
4, 31-36.
Gross, W. F., Curtain, M. E., & Moore, K. B. Appraisal of a milieu
therapy environment by treatment team and patients. Journal of
Clinical Psychology, 1971, 26, 541-545.
Gross, W. F., & DeRidder, L. M. Significant movement in comparatively
short-term counseling. Journal of Counseling Psychology, 1966,
13, 98-99.
Guertin, W. H., & Bailey, J. P. Introduction to Factor Analysis. Ann
Arbor: Edwards Brothers, 1970.
Haley, J. Strategies of Psychotherapy. New York: Grue and Stratton,
1963.

Hamrin, S. A., & Paulsen, B. P. Counseling Adolescents. Chicago:
Science Research Associates, 1950.
Hansen, J. C., Moore, G. D., & Carkhuff, R. R. The differential rela
tionships of objective and client perceptions of counseling.
Journal of Clinical Psychology, 1968, 24_, 244-246.
Hansen, J. C., & Warner, R. W., Jr. Review of research on practicum
supervision. Counselor Education and Supervision, 1971, 10,
261-272.
Hefele, T. J., & Hurst, M. W. InterpersonaI skill measurement: Pre
cision, validity and utility. The Counseling Psychologist, 1972,
3, 62-70.
Heller, K., Myer, R. A., & Kline, L. V. Interview behavior as a func
tion of standardized client roles. Journal of Consulting Psychol
ogy, 1963, 27, I 17-122.
Hill, G. E., & Green, D. A. The selection, preparation, and profession
alization of guidance and personnel workers. Review of Educational
Research, I960, 30, 115-130.
Hoffman, A. E. An analysis of counselor sub-roles. Journal of Coun
seling Psychology, 1959, 6, 61-67.
Hollenbeck, G. P. Conditions and outcomes in the student-parent rela
tionship. Journal of Consulting Psychology, 1965, ,29, 237-241.
Hough, J. B. The dogmatism factor in human relations training of pre
service teachers. Paper presented at the Educational Research
Association, Chicago, Illinois, February, 1965.
Howe, E. S., & Pope, B. The dimensionality of ratings of therapist
verbal responses. Journal of Consulting Psychology, 1961, 25,
296-300.
Hurst, M. W., & Shatkin, S. D. Relationship between standardized
admissions variables and certain interpersonal skills. Counselor
Education and Supervision, 1974, 14, 22-33.
Institute for Personality and Ability Testing, New prediction possi
bilities for vocationaI and educationaI counseI ing with the 16 PF,
Information Bulletin No. _9. Champaign, Illinois: IPAT, 1963.
Ivey, A. E. Microcounseling. Springfield, Illinois: Charles C.
Thomas, 1971.
Ivey, A. E., Normington, C. J., Miller, C. D., Morrill, W. H., & Haase,
R. F. Microcounseling and attending behavior: An approach to
pre-practicum counselor training. Journal of Counseling Psychology
Monograph, 1968, JJ5, 1-12.

119
Jackson, D. D. The study of the family: Family rules. Family Process,
1965, 12, 589-594.
Jennings, G. D. The relationship between perceptual characteristics
and effective advisinq of university housing para-professionaI
residence assistants. Doctoral dissertation, University of Florida,
1973.
Johnston, J. A. Practicum and on-the-job ratings of school counselors.
Personnel and Guidance Journa I, I 966, _45, 16-19.
Jones, A. J. Principles of Guidance and Personnel Work. New York:
McGraw-HiII, 1951.
Jones, L. K. The counselor evaluation rating scale: A valid criterion
of counselor effectiveness? Counselor Education and Supervision,
1974, J_4, I 12-1 16.
Kagan, N., & Krathwohl, D. (Eds.). Studies in Human Interaction: Inter
personal Process RecaI I St i mu Iated by V ideotape. U. S. Department
of Health, Education, and Welfare Project, No. 5-0800, December,
1967.
Kassera, W. J., & Sease, W. A. Personal change as a concomitant of
counselor education. Counselor Education and Supervision, 1970,
9, 208-21 I.
Kaul, T. J., Kaul, M. A., & Bednar, R. L. Counselor confrontation and
client depth of self-exploration. Journal of Counseling Psychology,
1973, 20, 132-136.
Keltz, J. W. The development and evaluation of a measure of counselor
effectiveness. Personnel and Guidance Journal, 1966, 44, 511-516.
Kemp, C. G. Influence of dogmatism on the training of counselors.
Journal of Counseling Psychology, 1962, 9_, 155-157.
Kemp, D. E. Correlates of the Whitehorn-Betz AB scale in a quasi-
therapeutic situation. Journal of Consulting Psychology, 1966,
30, 509-516.
Kerlinger, F. N. Foundations of Behavioral Research. New York: Holt
Rinehart and Winston, 1973.
Kiesler, D. J. Basic methodological issues implicit in psychotherapy
process research. American Journal of Psychotherapy, 1966, 20,
135-155.
Kilpatrick, F. P., & CantriI, H. Self-anchoring scaling: A measure of
individuals' unique reality worlds. Journal of Individual Psy
chology, I960, J_6, 158-174.

120
Klein, D. F., & Cleary, T. A. Platonic true scores and error in
psychiatric rating scales. Psychological Bulletin, 1967, 68, 77-80.
Knupfer, G., Jackson, D. D., & Krieger, G. Personality differences
between more and less competent psychotherapists as a function of
criteria of competence. Journal of Nerve and Mental Disorders,
1959, 129, 375-384.
Krumboltz, J. D. (Ed.) Revolution in Counselings: Imp Iications of
Behavioral Science. Boston: Houghton-M¡ffIin, 1966.
Kurtz, R. R., & Grummon, D. L. Different approaches to the measurement
of therapist empathy and their relationship to therapy outcomes.
Journal of Consulting and Clinical Psychology, 1972, 39, 106-115.
Lanning, W. L. A study of the relation between group and individual
counseling supervision and three relationship measures. Journal of
Counseling Psychology, 1971, J_8, 401-406.
Lanning, W. L., & Lemons, S. L. Another look at the factor structure of
the Barrett-Lennard Relationship Inventory. Measurement and Evalua
tion in Guidance, 1974, 6_, 228-231.
Lawlis, G. F., & Lu, E. Judgment of counseling process: Reliability,
agreement, and error. Psychological Bui let in, 1972, 78, 17-20.
Lesser, W. M. The relationship between counseling progress and empathic
understanding. Journal of Counseling Psychology, 1961, (3, 330-336.
Lin, T. T. Counseling relationship as a function of counselors self
confidence. Journal of Counseling Psychology, 1973, 20, 293-297.
Linden, J. D., Stone, S. C., & Shertzer, B. Development and evaluation
of an inventory for rating counseling. Personnel and Guidance
Journal. 1965, 44,267-276.
Lister, J. L. Reliability of the Helper Response Preference Inventory.
Unpublished mimeograph report, University of Florida, 1972.
Lister, J. L. School counseling: For better or for worse? Caadian
Counselor. 1970, 4_, 33-39.
Loesch, L. C. NonverbaIized feelings and perceptions of the counseling
relationship. Counselor Education and Supervision, 1975. 15. 105
113.
Loesch, L. C., & Rucker, B. B. Another view on the validity of the
Counselor Evaluation Rating Scale (submitted for publication).
Lou, M., & McNair, D. M. Methods relating to evaluation of therapeutic
outcome. In Gottschalk, L. A., & Auerback, A. H. (Eds.), Methods
of Research in Psychotherapy. New York: AppIeton-Century-Crofts,
1966.

121
MacGuffie, R. A., Janzen, F, V., & McPhee, W. M. The expression and
perception of feelings between student and supervisors in a practi-
cum setting. Counselor Education and Supervision, 1970, _9, 263-266.
McClain, E. W. Sixteen Personality Factor Questionnaire scores and
success in counseling. Journal of Counseling Psychology, 1968,
J_5, 492-496.
McDougall, W., & Reitan, H. The use of a peer rating technique in
appraising selected attributes of counselor trainees. Counselor
Education and Supervision, 1961, J_, 72-74.
Mason, J., & Blumberg, A. Perceived educational value of the classroom
and teacher-pupil interpersonal relationships. Journal of Secondary
Education, 1969, 44, 135-139.
McDaniel, W. Counselor selection: An evaluation of instruments. Coun
selor Education and Supervision, 1967, 6_, 142-144.
Mcllvaine, J. F. Coached clients as raters of counseling effectiveness.
Counselor Education and Supervision, 1972, 12, 123-129.
McNair, D. M., Callahan, D. M., & Lorr, M. Therapist "type" and patient
response to psychotherapy. Journal of Consulting Psychology, 1962,
26, 425-429.
McWhirter, J. J., & Marks, S. E. An investigation of the relationship
between the facilitative conditions and peer and group leader
ratings of perceived counseling effectiveness. JournaI of Clinical
Psychology, 1972, 28, I 16-117.
Mendoza, B. F. Predicting Counselor Effectiveness: A Multiple Regression
Approach. Doctoral dissertation, Western Michigan University,
1966.
Mills, D. H., & Zytowski, D. G. Helping relationship: A structural
analysis. Journal of Counseling Psychology, 1967, 14, 193-197.
Mueller, W. J., Gatsch, C. M., & Ralston, J. K. The prediction of
counselor interview behavior. Personnel and Guidance Journal,
1963, £l_> 513-517.
Munger, P. F., & Johnson, C. A. Changes in attitude associated with an
NDEA counseling and guidance institute. Personnel and Guidance
Journal, I960, 38, 751-753.
Myrick, R. D., & Kelly, F. D. A scale for evaluating practicum students
in counseling and supervision. Counselor Education and Supervision,
1971, J_0, 330-336.
Myrick, R. D., Kelly, F. D., and Wittmer, J. The Sixteen Personality
Factor Questionnaire as a predictor of counselor effectiveness.
Counselor Education and Supervision, 1972, I 1, 293-301.

122
National Vocational Guidance Association. Counselor Preparation. Wash
ington: The Association, 1949.
Nie, Norman H., Hull, C. Hadlai, Jenkins, Jean G,, Steinbrenner, Karen, &
Bent, Dale H. Statistical Package for the Social Sciences (2nd
ed.). New York: McGraw-Hill Book Company, 1975.
O'Hern, J. S., & Arbuckle, D. S. Sensitivity: A measurable concept?
Personnel and Guidance Journal, 1964, 4_2, 572-576.
Ohlsen, M. M. Final technical report on Institute for Elementary School
Counselors. Col lege of Education, University of Illinois, 1967.
Pal lone, N. J., & Grande, P. P. Counselor verbal mode, problem rele
vant communication and client rapport. Journal of Counseling
Psychology, 1965, J_2, 359-365.
Pare, D. D. Accurate Empathy Scale: Relative or Absolute? Doctoral
dissertation, University of Florida, 1970.
Patterson, C. H. A note on the evaluation of the effectiveness of
counseling and psychotherapy. Counselor Education and Supervision,
1964, 3, 129-131. (a)
Patterson, C. H. Client expectations and social conditioning. PersonneI
and Guidance Journal, 1958, 57, 136-138.
Patterson, C. H. Comments on client satisfaction, counselors, and the
counseling process. Personnel and Guidance Journal, 1959, 38, 25.
Patterson, C. H. Supervising students in the counseling practicum.
Journal of Counseling Psychology, 1964, J_l_, 47-55. (b)
Patterson, C. H. The selection of counselors. In J. M. Whiteley (Ed.),
Research in Counseling. Columbus, Ohio: Charles E. Merrill, 1967.
Peters, H. J., & Hansen, J. C. Counseling practicum: Bases for super
vision. Counselor Education and Supervision, 1963, 2, 82-85.
Pfeifle, H. H. Client perceptions of counselors with and without a
practicum. Counselor Education and Supervision, 1971, J_, 49-55.
Pierce, R., Carkhuff, R. R., & Berenson, B. G. The differential effects
of high and low functioning counselors upon counselors-in-training.
Journal of Clinical Psychology, 1967, 23, 212-215.
Pohlman, E. Changes in client preferences during counseling. PersonneI
and Guidance Journal, 1961, 40, 340-343,
Pohlman, E., & Robinson, F. P. Client reaction to some aspects of the
counseling situation. Personnel and Guidance Journal, I960, 38,
546-551.

123
Polmantier, P. C. The personality of the counselor. Vocational Guid
ance Quarterly, 1966, 16, 95-100.
Poole, A. Counselor judgment and counselees evaluation. Journal of
Counseling Psychology, 1957, 4_, 37-40.
Porter, E. H. Therapeutic Counseling. Boston: Houghton Mifflin Com
pany, 1950.
Purcell, F. E. Counseling assignments and efficiency. VocationaI
Guidance Quarterly, 1957, 5_, MI-113.
Reed, M. A study of client satisfaction as a criterion to evaluate
effective counseling. Unpublished master's thesis, University of
Utah, 1969.
Rapaport, A. The promise and pitfalls of information theory. In Buck-
ley, W. (Ed.), Modern Systems Research for the Behavioral Scien
tist. Chicago: Aldine, 1968.
Redfering, D. L. Changes in perception of the counselor: Self, ideal
self, and self as judged by peers. Counselor Education and Super
vision. 1973, _[3, 289-293.
Rickabaugh, K., Heaps, R. A., & Finley, R. Counselor comfort, counseling
climate, and client satisfaction: Client ratings and academic
improvement. Counselor Education and Supervision, 1972, JJ_, 219-
223.
Riese, H. C., & Stoner, W. G. The meaning of words: Teachers and
counselors. Counselor Education and Supervision, 1973, 13, 46-51.
Rogers, C. R. A tentative scale for the measurement of process psycho
therapy. In E. A. Rubinstein and M. B. Parloff (Eds.), Research
in Psychotherapy, Vo Iume I. Washington, D. C.: American Psycho
logical Association, 1959, 96-107.
Rogers, C. R. A process conception of psychotherapy. The American
Psychologist, 1958, J_3, 142-149.
Rogers, C. R. Psychotherapy todayor, where do we go from here?
American Journal of Psychotherapy, 1963, 17, 5-16.
Rogers, C. R. The necessary and sufficient conditions of therapeutic
personality change. Journal of Counseling Psychology, 1957, 21,
95-103.
Rogers, C. R., & Dymond, R. F. (Eds.). Psychotherapy and Personality
Change. Chicago: University of Chicago Press, 1954.
Rogers, C. R. Gendlin, E. T., Kiesler, D. J., & Truax, C. B. The Thera
peutic Relationship and its Impact: A_ Study of Psychotherapy with
Schizophrenics. Madison: University of Wisconsin Press, 1967.

124
Rosen, A. Client preferences: An overview of the Iiterature. Personnel
and Guidance Journal, 1967, 45, 785-789.
Rotter, J. C. The perceptuaI character istics of elementary, secondary,
and community college counselors. Doctoral dissertation, Wayne
State University, 1971.
Rushong, H. D. Present status and trends in the evaluation of counsel
ing. Educational and Psychological Measurement, 1953, 15, 418-430.
Russo, J. R., Kelz, J. W., & Hudson, G. R. Are good counselors open-
minded? Counselor Education and Supervision, 1964, _3, 74-77.
Satir, V. Conjoint Family Therapy. Palo Alto: Science and Behavior
Books, 1967.
Schmidt, G. D. Faculty perceptions of clergy in counselor education
programs. Counselor Education and Supervision, 1974, 13, 294-298.
Schroeder, P., & Dowse, E. Selection, function and assessment of
residence hall counselors. Personnel and Guidance Journal, 1968,
47_, 151-156.
Severinsen, K. N. Client expectations and perception of the counselor's
role and their relationship to client satisfaction. Journal of
Counseling Psychology, 1966, 13, 109-112.
Shapiro, J. G. Relationships between expert and neophyte ratings of
therapeutic conditions. Journal of Consulting and Clinical Psychol-
ogy, 1968, 32, 87-89.
Shoben, E. J. Research in counseling: A symposium; some problems in
establishing criteria of effectiveness. Personnel and Guidance
Journal 1953, 2U_, 284-294.
Snelbecker, G. E. Influence of therapeutic technigues on college
students' perceptions of therapists. Journal of Consulting Psychol
ogy, 1967, 31, 614-618.
Sonne, T. R., & Goldman, L. Preferences of authoratarian and eguali-
tarian personalities for client-centered and selectic counseling.
Journal of Counseling Psychology, 1957, £, 129-135.
Stablein, J. E. A critigue of counseling evaluation. Personnel and
Guidance Journal 1962, £l_, 66-68.
Steffi re, B., King, P., & Leafgren, F, Characteristics of counselors
judged effective by their peers. Journal of Counseling Psychology,
1962, 9, 335-340.
Stern, G. G., Stain, M. L., & Bloom, B. S. Methods in Personality
Assessment. Glencoe: Free Press, 1956.

125
Stone, A. R., Frank, J. D., Nash, E, H., & Imber, S, D, An intensive
five-year follow-up study of treated psychiatric patients. JournaI
of Nervous and Mental Disorders, 1961, 155, 410-421,
Stoughton, R. W. The preparation of counselors and personnel workers.
Review of Educational Research, 1957, 27, 174-185.
Stripling, R. 0., & Lister, J. L. Selection, preparation and profession
alization of a specialist. Review of Educational Research, 1963,
33. 171-178.
Sullivan, H. S. The Interpersonal Theory of Psychiatry. New York:
W. W. Norton, 1953.
Sunderlund, D, M., & Barker, E. N, The orientation of psychotherapists.
Journal of Consulting Psychology, 1962, 26, 403-409.
Thompson, A., & Miller, A. Steps toward outcome criteria in counseling
and psychotherapy, 1970, ERIC ED 070 010.
Thoresen, C. E. Relevance and research in counseling. Review of Educa
tional Research, 1969, 39, 264-282. (a)
Thoresen, C. E. The systems approach and counselor education: Basis
features and implications. Counselor Education and Supervision,
1969, 9, 17-31. (b)
Thornton, B. M. Dimensions of perceived relationship as related to
marital adjustment. Unpublished master's thesis, Auburn Univer
sity, I960.
Tomlinson, T. M., & Hart, J. F., Jr. A validation study of the process
scale. Journal of Consulting Psychology, 1962, 26, 74-78.
Tos, D. J., Frumkin, R. M., & Wilson, M. E. Intercorrelations of four
relationship components of the Barrett-Lennard Relationship Inven
tory. Psychological Reports, 1968, 23, 641-642.
Truax, C. B. An approach toward training for the aide-therapist:
Research and implications. Address: Symposium on non-traditionaI
preparation for helping relationships, American Psychological
Association Convention, Chicago, September 5, 1965.
Truax, C. B. Comparisons between control patients, therapy patients
receiving high conditions, and therapy patients perceiving low
conditions on measures of constructive personality change. Brief
Research Reports, Wisconsin Psychiatric Institute, University of
Wisconsin, 1962. (a)
Truax, C. B. The process of group psychotherapy. Relationships between
hypothesized therapeutic conditions and intrapersonal exploration.
Psychological Monographs, 1961, 75. (7, Whole No. 511).

126
Truax, C. B. Therapist empathy, warmth, and genuineness and patient
personality change in group psychotherapy: A comparison between
interaction unit measures, time sample measures, and patient
perception measures. Journal of Clinical Psychology, 1966, 22,
225-229.
Truax, C. B. The relationship between the patient's perceptions of the
level of therapeutic conditions offered in psychotherapy and con
structive personality change. Brief Research Reports, Wisconsin
Psychiatric Institute, University of Wisconsin, 1962. (b)
Truax, C, B., & Carkhuff, R. R. Concreteness: A neglected variable in
research in psychotherapy. Journal of Clinical Psychology, 1964,
20, 264-267.
Truax, C. B., & Carkhuff, R. R. Experimental manipulation of thera
peutic conditions. Journal of Consulting Psychology, 1965, 29,
119-124.
Truax, C. B., & Carkhuff, R. R. Toward Effective Counseling and Psycho
therapy. Chicago: Aldine Publishing Co., 1967.
Truax, C. B., & Mitchell, K. M. Research on certain therapist inter
personal skills in relation to process and outcome. In A. E.
Bergin and S. L. Garfield (Eds.) Handbook of psychotherapy and
behavior change: An empirical analysis. New York: John Wiley and
Sons, 1971.
Truax, C. B., Silber, L. D., & Wargo, D. G. Personality Changes and
Achievement in Therapeutic Training. Arkansas Rehabilitation
Research and Training Center, University of Arkansas, 1966.
Truax, C. B., Wargo, D. G., Carkhuff, R. R., & Tunnel, B. T., Jr.
Client perception of therapist empathy, warmth, and genuineness
and therapeutic outcome in group counseling with juvenile delin
quents. Unpublished manuscript, Arkansas Rehabilitation Research
and Training Center, University of Arkansas, 1966.
Truax, C. B., Wargo, D. G., Frank, J. D., Imber, S. D., Butle, B. C.,
Hoehn-Sarie, R., Nash, C. H., & Son, A. R. Therapist empathy,
genuineness, and warmth and patient therapeutic outcome. JournaI
of Consulting Psychology, 1966, 50, 395-401.
Truax, C. B., Wargo, D. G., Tunnel, B. T., Jr., & Glenn, A. W. Patient
perception of therapist empathy, warmth, and genuineness and thera
peutic outcome in outpatient group therapy. Unpublished manu
script, Arkansas Rehabilitation Research and Training Center,
University of Arkansas, 1966.
Van Der Veen, F. Dimensions of client and therapist behavior in rela
tion to outcome. Proceedinqs of the 73rd annua I convention of the
American Psychological Association. Washington, D. C.: American
Psychological Association, 1965, 279-280.

127
Walker, B. S., & Little, D. F. Factor analysis of the Barrett-Lennard
Relationship Inventory, Journal of Counseling Psychology, 1969,
16, 516-521.
Walker, A. M., Rabien, R. A., & Rogers, C, R, Development of a scale to
measure process changes in psychotherapy, JournaI of CI inicaI
Psychology, I960, 20, 79-85.
Walton, F. X., & Sweeney, T. J. Useful predictors of counseling effec
tiveness. Personnel and Guidance Journal, 1969, 48, 32-38.
Ward, G. R. Theoretical orientation influences the assessment of
counselor effectiveness. Counselor Education and Supervision, 1974,
14, 150-154.
Wasson, R. M. The Wisconsin Relationship Orientation Scale as a unique
variable in the assessment of applicants for counselor education.
Counselor Education and Supervision, 1965, 5_, 89-92.
Watley, D. J. Counselor predictive skill and rated counselor effec
tiveness. Personnel and Guidance Journal, 1967, 45, 579-584,
Wellman, F. E. A challenge and some problems. Proceedings of the First
Annual Invitational Conference for School Counselors, Ohio State
University, 1957.
Whitehorn, J. C., & Betz, J. A study of psychotherapeutic relation
ships between physicians and schizophrenic patients. American
Journal of Psychiatry, 1954, 3_, 321-331.
Whitehorn, J. C., & Betz, B. J. Further studies of the data as a cru
cial variable in the outcome of treatment with schizophrenic
patients. American Journal of Psychiatry, I960, I 17, 215-223.
Whiteley, J. M. Counselor Education. Review of Educational Research,
1969, 39, 173-187.
Wicas, E. A., & Mahan, T. W. Characteristics of counselors rated effec
tive by supervisors and peers. Counselor Education and Supervision,
1966, 6, 50-56.
Wittmer, J., & Lister, J. L. The graduate record examination, 16 PF
questionnaire, and counseling effectiveness. Counselor Education
and Supervision, 1971, J_0, 293.
Ziemelis, A. Effects of client preference and expectancy upon the initial
interview. Journal of Counseling Psychology, 1974, 21, 23-30.

BIOGRAPHICAL SKETCH
Sidney Leon Bradley was born in Pembroke, Georgia, on September
22, 1942, the second child of Samuel and Ida Mae Bradley. He graduated
from Bryan County High School in May, I960. He received the Bachelor
of Arts degree in theology from Berkshire Christian College in 1964.
That same year he married Barbara Ann Sherrill, and he and his wife
moved to Friendship, Maine, where he became minister of the Advent
Christian Church.
In 1967 he enrolled in the University of Georgia School of Social
Work and received the Master of Social Work degree in 1969 with a
specialization in psychiatric settings. From 1969 to 1972 he was
Director of the Children's Program at the Advent Christian Home. In
1972 he began graduate work in Counselor Education at the University of
Florida as a Graduate Council Fellow. In October, 1973, he received
a faculty appointment in the Department of Psychiatry, University of
Florida, after completing his doctoral practica in the Adult Outpatient
Clinic. He received the Specialist in Education degree in Counselor
Education at the University of Florida in June, 1974. He simultaneously
completed his doctoral internship in the Adult Outpatient Clinic with a
subspecialization in marriage and family counseling while serving there
as Director of Group and Family Therapy.
In February, 1976, Mr. Bradley returned to the Advent Christian
Home in Dowling Park, Florida, with his wife and two sons, Jonathan and
128

129
Stephen. Currently he is Director of the Child and Family Enrichment
Center and Vice President of the Advent Christian Home. He is a member
of the National Association of Social Workers and the Academy of Certi
fied Social Workers. He also holds clinical membership in the American
Association of Marriage and Family Counselors. Mr. Bradley holds
membership in the following honorary organizations: Delta Epsilon Chi,
Phi Kappa Phi, and Kappa Delta Pi.

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
P. Jo^£>h WiTtmer, Chairman
Professor of Counselor Education
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Paul G. Schauble
Professor of Psychology
This dissertation was submitted to the Graduate Faculty of the College
of Education and to the Graduate Council, and was accepted as partial
fulfillment of the requirements for the degree of Doctor of Philosophy.
December, 1976
Dean?Graduate School



98
42. She appreciates exactly how the things I experience feel to
me.
43. She approves of some things I do, and plainly disapproves of
others.
44. She is willing to express whatever is actually in her mind
with me, including any feelings about herself or about me.
45. She doesnt like me for myself.
46. At times she thinks that I feel a lot more strongly about a
particular thing than I really do.
47. Whether I am in good spirits or feeling upset does not make
her feel any more or less appreciative of me.
48. She is openly herself in our relationship.
49. I seem to irritate and bother her.
50. She does not realize how sensitive I am about some of the
things we discuss.
51. Whether the ideas and feelings I express are "good" or "bad"
seems to make no difference to her feeling toward me.
52. There are times when I feel that her outward response to me
is quite different from the way she feels underneath.
53. At times she feels contempt for me.
54. She understands me.
55. Sometimes I am more worthwhile in her eyes than I am at other
times.
56. I have not felt she tries to hide anything from herself that
she feels with me.
57. She is truly interested in me.
58. Her response to me is usually so fixed and automatic that I
don't really get through to her.
59. I don't think that anything I say or do really changes the
way she feels toward me.
60. What she says to me often gives a wrong impression of her whole
thought or feeling at the time.
61. She feels deep affection for me.
62. When I am hurt or upset she can recognize my feelings exactly,
without becoming upset herself.
63. What other people think of me does (or would, if she knew)
affect the ways she feels toward me.
64. I believe that she has feelings she does not tell me about
that are causing difficulty in our relationship.


38
3.
All supervisors had extensive counseling experience in a variety of
settings with a wide range of clients. Most of these supervisors
had completed a counseling internship of a full academic year in
addition to other supervised practica or clerkships before entry
into full-orbed professional practice. Every supervisor was a
faculty member in the Department of Counselor Education or had been
approved by the department. Each supervisor held an earned doctorate
in an appropriate field such as Counselor Education or Psychology.
All of the supervisors had conducted weekly individual supervision
with the counselor trainee being evaluated for a period of not less
than six weeks.
Clients. The clients in this study were persons seen by the coun
selor trainee in a helping relationship at the trainee's assigned
agency during the predetermined week of data collection. Each
client had to have been seen by the counselor trainee subject for
a minimum of two sessions of at least one half hour duration to be
included in the study.
Data Col lection
Data collection for this study occurred during the seventh and
eighth weeks of the Spring, 1976, quarter at the University of Florida.
During that time frame each counselor trainee subject was evaluated by
an individual supervisor. Only counselor trainees who received super
vision on the university campus or in the local community were included
in the sample. Also, evaluative ratings were made by each trainee's
individual supervisor only. Generally, the individual supervisor met
weekly with the trainee on a one-to-one basis for one hour per week.


Combs, A. W., & Soper, D. W. The perceptual organization of effective
counselors. Journal of Counseling Psychology, 1963, 10, 222-227.
Cornell, P. T. Factors influencing communication in counseling.
Doctoral dissertation, University of Missouri, 1955.
Cottle, W. C. Personal characteristics of counselors. Personnel and
Guidance JournaI, 1953, 31, 445-449.
Cottle, W. C., Lewis, W. W., Jr., & Penny, M. M. Personal characteris
tics of counselors. I I I An experimental scale. Journal of Coun
seling Psychology, 1954, J_, 74-77.
Cottle, W. C., & Lewis, W. W., Jr. Personality characteristics of
counselors: II Male counselor responses to MMPI and GZTS.
Journal of Counseling Psychology, 1954, J_, 27-30.
Cox, Rachel. Counselors and their work. Philadelphia: Archives
Publishing Company, 1945.
Cronbach, L. J. Essentials of psychological testing. New York:
Harper and Row, I960.
Davis, R. M., & OConner, J. B. Changes in counseling response prefer
ence among rehabilitation counseling interns. Counselor Educa
tion and Supervision, 1974, J_4, 40-46.
Dedrick, C. The relationship between perceptual characteristics and
effective teaching at the junior college level. Doctoral disser
tation, University of Florida, 1972.
Demos, G. D., & Zuwaylif, F. H. Characteristics of effective counselors.
Counselor Education and Supervision, 1966, 5_, 163-165.
Dickman, J. The perceptual organization of person oriented versus task
oriented student nurses. Doctoral dissertation, University of
Florida, 1967.
Dilley, J. Supervisory ratings of counselor trainees in a simulated
work setting as compared with peer and instructor ratings of the
same trainees in an academic setting. Counselor Education and
Supervision, 1964, ~b_, 70-73.
Dole, A. A. The prediction of effectiveness in school counseling.
Journal of Counseling Psychology, 1964, JJ_, I 12-121.
Donnan, H. H., Harlan, G. E., & Thompson, S. A. Counselor personality
and level of functioning as perceived by counselees. Journal of
Counseling Psychology, 1969, J_6, 482-485.
Doy le, E. J. The relationship between coI Iege teaching effectiveness
and inferred characteristics of the adeguate personality. Doctora I
dissertation, University of Northern Colorado, 1969.


35
of acceptance and continue to be regarded as one of the most valid
criterion measures. Client ratings have yielded more mixed results, but
the weight of ev
ratings are used
is evidence that
dence supports their use. When supervisor and client
in conjunction, as they have been in this study, there
each strengthens the credibility of the other.


70
instruments, the Barrett-Lennard Relationship Inventory (BLRI) and the
Counselor Evaluation Rating Scale (CERS), and for the predictor instru
ment, the Helper Response Preference Inventory (HRPI). There are 204
positive correlations and 120 negative correlations.
Significant in this matrix are the consistently negative correla
tions between BLRI scores and CERS scores. The mean correlation for all
correlations between the two criterion instruments is -.21. All correla
tions are negative except for the correlation between the BLRI Level of
Regard score and the CERS Supervision score which is a positive, though
ngligible, .02.
Canonical correlations were computed between subscale scores of the
BLRI and subscale scores of the CERS. Similar operations were performed
relating the HRPI variables to each set of the criterion instrument
subscales. In all cases the canonical correlations failed to reach the
.05 level of significance.
The relationships of each of the ten HRPI factors to the combined
scores of the BLRI and also to the combined scores of the CERS are
presented in Table 9. The statistic was derived by averaging the
correlations of the predictor HRPI factor with all subscales and total
score of each criterion instrument. The first column reveals the mean
correlation of the specified HRPI factor with the five subscores of the
BLRI. The second column shows the mean correlation of the specified
HRPI factor with the three subscores of the CERS. Notable are the seven
correlations, either positive or negative, above .15. Using this global
measure the .20 correlation of Factor E with the CERS scores and the .21
correlation of Factor C with the CERS scores seem worthy of more examina
tion. Likewise the correlations of Factors H and J with the BLRI scores


28
effectiveness in terms of demonstrable client outcome criteria. In
response, the Carkhuff model has endeavored to derive a coherent system
of selection and training based upon their conclusions from outcome
research. Nevertheless there remain the ubiquitous problems of effec
tiveness criteria, plausible indicators of therapeutic gain, consensual
counseling goals, and quantifying the variability in clients, settings,
and problems. While the ultimate goal has not been attained, the
impression is that greater definitional and procedural clarity have
evolved from the continual grappling with the problem of counselor
effectiveness criteria.
Counselor Effectiveness Ratings
The widespread use of global ratings of effectiveness as criteria
in counseling research has been previously mentioned. In their review
of the effectiveness literature, Walton and Sweeney (1969) point out
the broad base of acceptance rating scales have attained. Since the
original studies by Rogers and Dymond (1954) and the emergence of the
"Process Scale" conceptualized (1958) and published by Rogers (1959),
there have arisen a host of researchers and research instruments using
y I t. J y l 1 f J I S '. i -
a ratings format.
In addition to the Carkhuff scales, already cited, similar rating
scales include the Self-Anchoring Scale of Counseling Effectiveness
(Kilpatrick & Cantril, i960), the Communication Rating Scale (Brams,
1961), the Counselor Potential Scale (Dole, 1964), the Wisconsis Rela
tionship Orientation Scale (Wasson, 1965), the Counselor Evaluation
Inventory (Linden, Stone, & Schertzer, 1965), the Counselor Evaluation
Form (FauIkenberry, 1968), the Counselor Practicum Evaluation Form


12
in this area. Walton and Sweeney (1969) state that in their review
they found that "results of work utilizing this instrument have rather
consistently supported its capability for distinguishing between coun
selors on opposite ends of the effectiveness continuum" (p. 34).
Steffi re, King, and Leafgren (1962) found, among counselor trainees
selected as having more counseling potential by their peers in an MDEA
Guidance and Counseling Institute, significantly less dogmatism than,
among those trainees not chosen. Cahoon, in 1962, also reported that he
had found significantly less dogmatism in superior counselors. These
findings have been supported by Kemp's finding (1962) that low scores in
dogmatism on the Dogmatism Scale were related to openness in an actual
counseling situation. Still further corroboration came from the conclu
sion by Russo, Kelz, and Hudson (1964) that openmindedness is indeed an
important quality in a counselor after they found 12 items on the
Dogmatism Scale which distinguished significantly between counselors
judged most and least effective. Furthermore, in studies that did not
use the Dogmatism Scale but measured a closely related factor, tolerance
for ambiguity, both Brams (1961) and McDaniel (1967) found a relation
ship between this factor and high ratings of effectiveness.
Virtually a myriad of other personality measures have been tried in
attempts to isolate personality patterns which appear to coincide with
effective counseling. Some illustrative efforts have included score
profiles from the Miller Analogies Test and the MMPI (Abeles, 1958),
the MMPI alone (Brams, 1961; Foley & Proff, 1965), the Strong Vocational
Interest Blank (Stefflre, King, & Leafgren, 1962), the Kuder Preference
Record (Blocher, 1963), the Wisconsin Relationship Orientation Scale
(Wasson, 1965), the Al Iport-Vernon-Lindzey Scale of Values (Demos &


97
.16.
.17.
18.
.19.
20.
21
22.
23.
24.
25.
26.
27.
28.
29.
30.
.31 .
32.
33.
34.
.35.
36.
37.
38.
39.
40.
41 .
It makes her uneasy when I ask or talk about certain things.
She is indifferent to me.
She usually senses or realizes what I am feeling.
She wants me to be a particular kind of person.
I nearly always feel that what she says expresses exactly what
she is feeling and thinking as she says it.
She finds me rather dull and uninteresting.
Her own attitudes toward some of the things I do or say pre
vent her from understanding me.
I can (or could) be openly critical or appreciative of her
without really making her feel any differently about me.
She wants me to think that she I ikes me or understands me
more than she really does.
She cares for me.
Sometimes she thinks that J_ feel a certain way, because thats
the way she fee Is.
She likes certain things about me, and there are other things
she does not like.
She does not avoid anything that is important for our rela-
tionship.
I feel that she disapproves of me.
She realizes what I mean even when I have difficulty in saying
it.
Her attitude toward me stays the same: she is not pleased
with me sometimes and critical or disappointed at other times.
Sometimes she is not at all comfortable but we go on, out
wardly ignoring it.
She just tolerates me.
She usually understands the whole of what I mean.
If I show that I am angry with her she becomes hurt or angry
with me, too.
She expresses her true impressions and feelings with me.
She is friendly and warm with me.
She just takes no notice of some things that I think or feel.
How much she likes or dislikes me is not altered by anything
that I tell her about myself.
At times I sense that she is not aware of what she is really
feeling with me.
I feel that she really values me.


100
16. Works well with other professional personnel (e.g. teachers,
counselors, etc.)
17. Can be spontaneous in counseling, yet behavior is relevant.
18. Lacks self-confidence in establishing counseling relation
ships.
19. Can explain what is involved in counseling and discuss
intelligently its objectives.
20. Is open to self-examination during supervision.
21. Can express thoughts and feelings clearly in counseling.
22. Verbal behavior in counseling is appropriately flexible and
varied, according to the situation.
23. Lacks basic knowledge of fundamental counseling principles
and methodology.
24. Participates actively and willingly in supervisory sessions.
25. Is indifferent to personal development and professional
growth.
26. Applies a consistent rationale of human behavior to counseling.
27. Can be recommended for a counseling position without reserva
tion.
Recommended Grade
Comments:


66
TABLE 7
MEANS, STANDARD DEVIATIONS, AND RANGES OF SCORES ON THE
HELPER RESPONSE PREFERENCE INVENTORY (N=46)
Factors
Means
Deviations
Ranqe
A.
(Empathic Understanding-
Probing)
24.22
4. 1 1
17-27
B.
(Confrontation-False Reas
surance)
16.26
3.21
8-21
C.
(Specificity-Ambiguity)
18.11
2.68
10-20
D.
(Analytical 1nterest-Va1i-
dation)
4.13
2.60
0-10
E.
(Clarificat ion-Superficia1ity)
13.57
2.35
8-16
F.
(Focused Probing-Premature
Judgements)
8.02
2.68
l-l 1
G.
(Tentative Formulations-
Avoidance of Emotion)
5.76
1.75
2-7
H.
(Requesting-Giv¡ng Information)
4.61
1.48
3-8
1.
(1nterna1-Externa 1 Frame of
Reference)
5.17
2.03
0-8
J.
(Frank Feedback-1ncongruence)
8.52
1.55
5-10
the upper half of possible scores. Three factors (D, F, and I) yielded
scores that covered almost the entire possible range. Factors B, G,
and H produced scores somewhere between these two extreme groups. No
other discernible pattern appeared to emerge from this set of data.
Relationships Among the Three Instruments
Table 8 presents the intercorrelation matrix for the two criterion


CHAPTER V
DISCUSSION AND SUMMARY
The review of research on counseling effectiveness in Chapter II
demonstrated the varied and often circuitous routes to building a con
sistent model for selecting and training effective counselors. Never
theless, the centrality and interrelatedness of the counselors person
ality and communication style emerge consistently across research as fun
damental to the helping process. The literature review also suggested
that counselors may have characteristic manners of responding verbally
to clients. The propensity toward preferred responses is probably rooted
in both habit and value structure and may crucially shape the forma
tion of the therapeutic relationship as well as the outcome of the
helping process.
This study was designed to investigate the relationship of response
choices on the Helper Response Preference Inventory by counselor
trainees to client perceptions of the relationship and supervisor
evaluations. Client perceptions were assessed through the Barrett-
Lennard Relationship Inventory (BLRI) and supervisor ratings were
collected by means of the Counselor Evaluation Rating Scale (CERS).
It was hypothesized, first of all, that a relationship would be
found between counselor trainee response preferences on the HRPI and
client evaluations on the BLRI. Second, it was hypothesized that a
relationship would be found between response preferences on the HRPI
79


23
junior college teachers (Dedrick, 1972), nurses (Dickman, 1967), college
professors (Doyle, 1969), elementary school teachers (Gooding, 1974),
residence assistants (Jennings, 1973), and counselors (Rotter, 1971).
The overall impression is that this substantial body of research suggests
a potentially rewarding means of identifying helpers with relatively
higher levels of helping ability.
The Carkhuff-Truax Therapeutic Conditions Model
The model of counselor effectiveness research which was popularized
by Truax and Carkhuff, and currently is often identified with the latter,
was formulated primarily out of the early research efforts of the pair
(Truax & Carkhuff, 1967). Their joint investigations were directed at
specifying the primary core of interpersonaI facilitative factors which
relate to favorable outcome in counseling and therapy. After extensive
research, the two concluded that the core dimensions of accurate empathy,
nonpossessive warmth, genuineness, and concreteness account for between
one-third to one-half the variability in client outcome (Truax, 1961;
Truax & Carkhuff, 1966). Counselor trainees who offer their clients
high levels of these core dimensions have been found to consistently
elicit beneficial therapeutic gain (Berenson, Carkhuff, & Myrus, 1966;
Carkhuff & Truax, 1965; Truax & Mitchell, 1971).
The work of Truax and Carkhuff is clearly an outgrowth of the
Rogers' formulations. Truax, a student of Rogers, developed a "Tenta
tive Scale for the Measurement of Depth of Self Exploration" along with
other scales rooted in the theoretical hypothesis of "necessary and
sufficient conditions" of therapeutic personality change proclaimed by
Rogers (1957). Much of the plethora of research summarized by Truax


Carkhuff, R. R., Piaget, G., & Pierce, R. The development of skills
in interpersonal functioning. Counselor Education and Supervision,
1968, 7, 102-106.
Carkhuff, R. R., & Truax, C. B. Training in counseling and psycho
therapy: An evaluation of an integrated didactic and experiential
approach. Journal of Consulting Psychology, 1965, 29, 333-336.
Carkhuff, R. R., & Truax, C. B. Toward explaining success and failure
in interpersonaI learning experiences. Personnel and Guidance
Journal, 1966, 44, 723-727.
Carson, R. C., Harden, J. A., & Shows, W. D. A-B distinction and
behavior in quasi-therapeutic situations. Journal of Consulting
Psychology, 1964, 28, 426-433.
Cartwright, R. D., & Lerner, B. Empathy, need to change, and improve
ment with psychotherapy. Journal of Consulting Psychology, 1963,
27, 138-144.
Cartwright, R. D., & Vogel, J. L. A comparison of changes in psycho
neurotic patients during matched periods of therapy and no therapy.
Journal of Consulting Psychology, I960, 24, 121-127.
Cash, W. L., Jr., & Munger, P. F. Counselors and their preparation.
Review of EducationaI Research, I 966, 36, 256-263.
Cattell, J. M. Statistics of American psychologists. American Journal
of Psychologists, 1903, J_4, 310-328.
Chap line, E. B. A study of cognitive complexity, selected demographic
variabIes and counseIinq behavior of counselor-trainees. (Doc
toral dissertation, Temple University) Ann Arbor, Mich.: Univer
sity Microfilms, 1964. No. 64-9329.
Chinsky, J. M., & Rappaport, J. Brief critique of the meaning and
reliability of "accurate empathy" ratings. Psycho logical Bulletin,
1970, 73, 379-382.
Colby, K. M. Computer simulation of a neurotic process. In Tomkins,
S. S. & Messick, S. (Eds.) Computer Simulation of Personality.
New York: John Wiley, 1963.
Combs, A. W. A perceptual view of "helpers" in personality theory and
counseling practice. First Annual Conference on Personality
Theory and Counseling Practice. Gainesville: University of
Florida Press, 1961.
Combs, A. W., Avila, D. W., & Purkey, W. W. Helping relationships:
Basic concepts for the helping professions. Boston: AI Iyn-Bacon,
1971. '
Combs, A. W., & Snygg, D. Individual behavior. New York: Harper and
Row, 1959.


25
and his associates, it seems appropriate to elucidate how he concep
tualizes the integration of the facilitative dimensions into the thera
peutic process. Two of his primary propositions and their corollaries
follow. These bold assertions are supported by research evidence in the
book Helping and Human Relations, Volume I (1969) from which they were
extracted.
Proposition I. The degree to which the helping person offers
high levels of facilitative conditions in response to the
expressions of the person seeking help is related directly
to the degree to which the person seeking help engages in
processes to constructive change or gain.
Corollary I. The degree to which the helping person offers
high levels of empathic understanding of the helpees
world is related directly to the degree to which the helpee
is able to understand himself and others.
Corollary II. The degree to which the helping person communi
cates high levels of respect and warmth for the helpee and
his world is related directly to the degree to which the
helpee is able to respect and direct warm feelings toward
himself and others.
Corollary III. The degree to which the helper is helpful in
guiding the exploration to specific feelings and content is
related directly to the degree to which the helpee is able
to make concrete his own problem areas.
Corollary IV. The degree to which the helper is responsively
genuine in his relationship with the helpee is related to the
degree to which the helpee is able to be responsively genuine
in his relationship with himself and others.
Proposition II. The degree to which the helping person
initiates action-oriented dimensions in a helping relation
ship is directly related to the degree to which the person
seeking help engages in processes that lead to constructive
change or growth.
Corollary I. The degree to which the helper can be freely,
spontaneously and deeply himself, including the disclosing
of significant information about himself when appropriate,
is directly related to the degree to which the helpee is
able to be genuine and self-disclosing in appropriate
relationships.
Corollary II. The degree to which the helper actively


14
to counseling effectiveness is mystifying. It is enigmatic why results
overall have been either contradictory or confusing. One criticism of
this entire approach came very early from Stern, Stein, and Bloom (1956)
They felt the inconsistency eventuated from the inadequacy of the trait
factor philosophy in which the approach was rooted. They called for
attempts to identify meaningful personality traits related to counseling
behavior only after predictions have been derived from we I I-art¡cu Iated
theories.
Other observers take a different perspective and view the problem
as an artifact of the "effectiveness" criteria invoked. The profession
of counseling has not attained a consensus on the goals of counseling,
the differential counseling methods appropriate to specific problems,
or standard indicators of client progress. Therefore, the elusive con
cept of "counselor effectiveness" may well be relative not only to the
uniqueness of the counselor, but also to the setting, the problem, and
the client (Stefflre, King, & Leafgren, 1962).
Steffi re, King, and Leafgren (1962) pointed out that we are unable,
at our present knowledge level, to reach consensus about how to identify
or even define the effective counselor. Nevertheless, societal pressure
urges the profession to educate "as if" the "effective counselor" was a
definable entity. Generally, ratings of supervisors or judgment by pro
fessional practitioners has been used to differentiate levels of skill,
but client ratings have gained in popularity and are explored in depth
in the next section of this chapter. However, the ratings utilized have
proliferated in effectiveness studies to include peer judgments, self-
ratings, etc., which has added to the profusion of confusion.
Researchers have attempted at times to look at behavioral criteria


59
optimum combination of predictor variables as well as for any separate
predictor variable which was, by itself, a significant predictor of
either of the dependent variables.
While not central to the purposes of this study, canonical corre
lations between all ten HRPI variables and the four Barrett-Lennard
Relationship Inventory subscore variables were computed to establish
the maximum correlation possible between the two sets of variables and
to estimate the variance shared by the two composites of variables. The
same procedure was used with the three Counselor Evaluation Rating
Scale scores and the ten HRPI variables. Finally, canonical correla
tions were computed for the four Barrett-Lennard Relationship subscore
variables and the three Counselor Evaluation Rating Scale score varia
bles. An intercorrelation matrix of all the variables in this study
is presented on page 67. It was anticipated that an intercorrelation
matrix could help clarify the results and could possibly explain the
presence or absence of statistically significant relationships among the
variables under scrutiny.


APPENDIX D
RELATIONSHIP INVENTORYFORM OS-F-64
Below are listed a variety of ways that one person may feel or behave in
relation to another person.
Please consider each statement with reference to your present relation-
ship
with your
Mark each statement in the left margin, according to how strongly you
feel that it is true, or not true, in this relationship. Please mark
every one. Write in +3, +2, +1, or -1. -2, -3. to stand for the follow-
ing answers:
+3:
Yes, 1 strongly feel that it
is true.
-1 :
No, 1 feel that it is probably
untrue or more untrue than true.
+2:
Yes, 1 feel it is true.
-2:
No, 1 feel it is not true.
+ 1 :
Yes, 1 feel that it is
probably true or more true
than untrue.
-3:
No, 1 strongly feel that it is
not true.
1. She respects me as a person.
2. She wants to understand how I see things.
3. Her interest in me depends on the things I say or do.
4. She is comfortable and at ease in our relationship.
5. She feels a true liking for me.
6. She may understand my words but she does not see the way I
feel.
7. Whether I am feeling happy or unhappy with myself makes no
real difference to the way she feels about me.
8. I feel that she puts on a role or front with me.
9. She is impatient with me.
10. She nearly always knows exactly what I mean.
11. Depending on my behavior, she has a better opinion of me
sometimes than she has at other times.
12. I feel that she is real and genuine with me.
13. I feel appreciated by her.
14. She looks at what I do from her own point of view.
15. Her feeling toward me doesn't depend on how I feel toward her.
96


32
Some appraisal of the client's reaction to the counselor and
to counseling should be obtained before we can say that we
have any comprehensive understanding of who makes a good
counselor ... a successful practitioner, among other things,
is one who elicits favorable reactions from the recipients
of his services. (Grigg & Goodstein, 1957, p. 31)
Indirect support for counselor ratings may be drawn from studies
which demonstrate significant positive correlations between client and
supervisor rankings of counselor effectiveness (Brown & Cannaday, 1969).
In several studies using the Counselor Evaluation Inventory (CEI) devised
by Linden, Stone and Shertzer (1965) high correlations between client
and supervisor ratings lend support to the hypothesis that clients and
supervisors tend to evaluate a similar set of criteria (Anderson &
Anderson, 1962; CorrelI, 1955; Poole, 1957). In research involving the
same instrument, Rickabaugh, Heaps, and Finley (1972) discovered a
significant relationship between clients' perceptions of counselor com
fort and client academic grade improvement, another frequently used
measure of effective counseling outcome. Pfeifle (1971) also found that
clients rated counselors with practicum experience significantly higher
than counselors without practical experience when ranked on the CEI.
Client ratings have often been an integral component in the pre
viously discussed research on the therapeutic conditions of empathy,
warmth, genuineness, respect, and concreteness offered by the counselor
(Truax & Carkhuff, 1967). One of the pioneers in using client percep
tions to measure the therapeutic conditions was Barrett-Lennard (1962).
An extensive analysis of the characteristics and uses of his instrument
follows in the section of the next chapter on "Instrumentation." The
use of his instrument and of similar relationship inventories has yielded
positive results with client samples ranging from juvenile delinquents


53
They concluded, from the correlation range between subscales of from
.31 to .58, that their "data appear to support Barrett-Lennard's con
clusion that the four subscales are relatively independent" (p. 642).
In 1974 Lanning and Lemons administered the OS form of the Rela
tionship Inventory on three separate occasions to compile data for a
pri ncipa I-components analysis and an orthogonal rotation of the factor
matrix by the varimax procedure. They found two nearly orthogonal
factors with Level of Regard, Empathy, and Congruence loading on a
component that accounts for from 62 to 72 percent of the total variance.
Unconditionality of Regard loads predominantly on the other factor which
accounts for 16 to 21 percent of the total variance. The authors felt
their study provided some support for the earlier findings of Mills and
Zytowski. Lanning and Lemons concluded:
The major component is probably a measure of the overall
satisfaction with the relationship, contrary to earlier claims
made about the separateness of the subtests. This indicates
that there may be one predominant characteristic across all
of the relationships measured rather than four separate ones.
(1974, p. 230)
Since there appeared to be somewhat inconclusive empirical support
for or against a general relationship satisfaction factor, this study
employed the total score of the Barrett-Lennard Relationship Inventory
and each sub-score as criterion measures based upon client ratings of
perceived counseling effectiveness. Both the male and female forms of
the OS-64 revision (see Appendix) were utilized in this study.
Counselor Evaluation Rating Scale (CERS)
The Counselor Evaluation Rating Scale (see Appendix) which was
used in this study as a criterion measure of supervisor rating, com
prises 27 items that enable a supervisor to rate a counselor trainee's


85
A different pair of HRPI factors predicted two CERS scores. With
the supervisor criterion no demographic variable emerged as a predictor.
Also with the CERS, self disclosure did not appear to be meaningful
in the prediction process. The supervisor ratings were related to
clarification. Perhaps the supervisors rated positively the ability of
the counselor trainee to cut through superficial verbiage to highlight
implied meanings and lead the interchange into sharper focus. Clarifi
cation was the only predictor of the Total score and by far the best
predictor of the Counseling score. The supervisors evidenced that they
value the trainees skill in leading the client into deeper strata of
self understanding. This theme is supported by the only other predictor
of a CERS score. The Counseling score was also predicted by responses
that validate the client's experience. The implication is, "Yes, thats
the way you really are." The result is a very positive, affirming way
of developing the client's repertoire of personal insights.
Thus, in an overall perspective it can be said that the HRPI factor
scores predict both the BLRI and the CERS in divergent ways. For the
client data the response choices that favor frank self disclosure are
significantly predictive. For supervisor ratings it is those response
preferences that stress development of client self understanding which
predict at a meaningful level.
AppIication
The pertinent question that attends a study of this nature arose
in the introduction in Chapter I, persisted throughout Chapter II, and
in essence permeates this research. The question is whether the predic
tor instrument under investigation can or should be used for counselor


TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS ¡
TABLE OF CONTENTS iv
LIST OF TABLES vi
ABSTRACT v i i
CHAPTER I INTRODUCTION I
Purpose of the Study I
Rationale for the Study I
Statement of the Problem 4
Hypotheses 5
Definition of Terms 5
CHAPTER II REVIEW OF RELATED RESEARCH 7
Approaches to Counselor Effectiveness 7
The Speculative-Philosophical Model 8
The Psychometric Model 10
The Criteria Crisis in Effectiveness Model Building ... 13
The Communications Analysis Model 17
The Perceptual Model 19
The Carkhuff-Truax Therapeutic Conditions Model 23
Summary 27
Counselor Effectiveness Ratings 28
Supervisor Ratings 30
Client Ratings 31
Summary 34
CHAPTER I I I METHODOLOGY 36
Description of the Sample 36
Data Col lection 38
Limitations 41
Instrumentation 42
Helper Response Preference Inventory (HRPI) 42
Barrett-Lennard Relationship Inventory (BLRI) 46
Counselor Evaluation Rating Scale (CERS) 53
Analysis of Data 57


CHAPTER IV RESULTS OF THE STUDY
60
Introduction 60
Descriptive Data of the Sample 60
The Counselor Trainees 60
Supervisors 63
Cl ients 63
Client Evaluation Scores on the Counselor Trainees 63
Supervisor Evaluation Scores on the Counselor Trainees .... 64
Scores of the Counselor Trainees on the Predictor
Instrument 65
Relationships among the Three Instruments 66
Testing the Hypotheses 72
Test of the First Major Hypothesis 72
Test of the Second Major Hypothesis 74
CHAPTER V DISCUSSION AND SUMMARY
Summary of the Results 80
Practical Limitations 81
Interpretation and Discussion of Findings 83
Appl ¡cation 85
Suggestions for Further Research 87
Conclusions of the Investigation 88
APPENDICES 90
A Counselor Trainee Demographic Data Sheet 91
B Cover Letter to the Client 92
C Relationship Inventory Form OS-M-64 93
D Relationship Inventory Form 0S-F-64 96
E Counselor Evaluation Rating Scale 99
F Helper Response Preference Inventory 101
BIBLIOGRAPHY NO
BIOGRAPHICAL SKETCH 128
v


127
Walker, B. S., & Little, D. F. Factor analysis of the Barrett-Lennard
Relationship Inventory, Journal of Counseling Psychology, 1969,
16, 516-521.
Walker, A. M., Rabien, R. A., & Rogers, C, R, Development of a scale to
measure process changes in psychotherapy, JournaI of CI inicaI
Psychology, I960, 20, 79-85.
Walton, F. X., & Sweeney, T. J. Useful predictors of counseling effec
tiveness. Personnel and Guidance Journal, 1969, 48, 32-38.
Ward, G. R. Theoretical orientation influences the assessment of
counselor effectiveness. Counselor Education and Supervision, 1974,
14, 150-154.
Wasson, R. M. The Wisconsin Relationship Orientation Scale as a unique
variable in the assessment of applicants for counselor education.
Counselor Education and Supervision, 1965, 5_, 89-92.
Watley, D. J. Counselor predictive skill and rated counselor effec
tiveness. Personnel and Guidance Journal, 1967, 45, 579-584,
Wellman, F. E. A challenge and some problems. Proceedings of the First
Annual Invitational Conference for School Counselors, Ohio State
University, 1957.
Whitehorn, J. C., & Betz, J. A study of psychotherapeutic relation
ships between physicians and schizophrenic patients. American
Journal of Psychiatry, 1954, 3_, 321-331.
Whitehorn, J. C., & Betz, B. J. Further studies of the data as a cru
cial variable in the outcome of treatment with schizophrenic
patients. American Journal of Psychiatry, I960, I 17, 215-223.
Whiteley, J. M. Counselor Education. Review of Educational Research,
1969, 39, 173-187.
Wicas, E. A., & Mahan, T. W. Characteristics of counselors rated effec
tive by supervisors and peers. Counselor Education and Supervision,
1966, 6, 50-56.
Wittmer, J., & Lister, J. L. The graduate record examination, 16 PF
questionnaire, and counseling effectiveness. Counselor Education
and Supervision, 1971, J_0, 293.
Ziemelis, A. Effects of client preference and expectancy upon the initial
interview. Journal of Counseling Psychology, 1974, 21, 23-30.


64
TABLE 5
MEANS, STANDARD DEVIATIONS, AND RANGES OF SCORES ON THE
BARRETT-LENNARD RELATIONSHIP INVENTORY (N=46)
Relationship
Dimensions
Means
Standard
Deviations
Range
Level of Regard
35.83
10.27
2-48
Empathic Understanding
30.54
12.84
1-48
Unconditionality of Regard
21.04
12.02
1-48
Congruence
33.52
11.50
1-48
Total Score
118.35
42.91
10-192
Of interest is the fact that the means of the sub scales all fall
in the low to mid thirties except for the Unconditionality of Regard
score which dips to 21.04. Nevertheless, there is no appreciable
difference in the standard deviations of all four subscores which fall
in the range of 10.27 to 12.84. The Total score had a mean of 118.35
and a standard deviation of 42.91.
Supervisor Evaluation Scores on the Counselor Trainees
The individual supervisor ratings on the Counselor Evaluation
Rating Scale are presented in Table 6. The summary shows the means,
standard deviations, and ranges of the scores for all 46 counselor
trainee subjects. Ranges on both the Counseling and Supervision sub
scales are quite broad, 33 to 91 and 38 to 91, respectively. The Total
score reflects a 91 to 189 point range.
The means of the two subscales, 81.33 for Counseling and 84.22 for


18
Munger & Johnson, I960). A paradoxical finding, however, has been dis
covered among rehabilitation counseling trainees where preferences
are decidedly in favor of responses that inform, question, probe and
diagnose (Davis & O'Conner, 1974; Wittmer & Lister, 1971).
In his doctoral dissertation Chap line (1964) related level of cog
nitive complexity in counselors to preferences for understanding or
evaluative responses. He administered the Construct Repertory Test to
his sample of 163 graduate students in counseling and guidance from the
departments of psychology and education to measure cognitive complexity.
In addition to gathering demographic data, he had the subjects respond
in writing to an experimental tape recording of client statements.
Using Porter's original categories (1950), he found that subjects who
measured high on cognitive complexity were also high on evaluative
responses while the low cognitive complexity scorers were high on
understanding responses. The education students tended to prefer
evaluative responses and the psychology students leaned toward under
standing responses at a higher-than-chance level. On the other hand,
response preferences were unrelated to sex, age, level of training,
previous personal counseling, teaching experience, or even experience in
counseling. Freedman, Antenem, and Lister (1967) in a related study
found a strong predictable relationship between verbal response patterns,
as measured by Porter's Scales, and counselor personality characteris
tics on the California Psychological Inventory. The response-analysis
of their 15-minute roleplayed interviews was suggested as a useful
procedure for student selection and progress evaluation.
Grigg and Goodstein (1959) found that clients who saw their coun
selors as verbally active were much more likely to report a favorable


94
J7. He is indifferent to me.
18. He usually senses or realizes what I am feeling.
19. He wants me to be a particular kind of person.
20. I nearly always feel that what he says expresses exactly what
he is feeling and thinking as he says it,
_2I. He finds me rather dull and uninteresting.
22. His own attitudes toward some of the things I do or say
prevent him from understanding me.
23. I can (or could) be openly critical or appreciative of him
without really making him feel any differently about me.
24. He wants me to think that he likes me or understands me more
than he really does.
25. He cares for me.
26. Sometimes he thinks that J_ feel a certain way, because that's
the way he_ feels.
27. He likes certain things about me, and there are other things
he does not like.
28. He does not avoid anything that is important for our relation
ship.
29. I feel that he disapproves of me.
30. He realizes what I mean even when I have difficulty in saying
it.
31. His attitude toward me stays the same: he is not pleased
with me sometimes and critical or disappointed at other times.
32. Sometimes he is not at all comfortable but we go on, outwardly
ignoring it.
33. He just tolerates me.
J54. He usually understands the whole of what I mean.
35. If I show that I am angry with him he becomes hurt or angry
with me, too.
36. He expresses his true impressions and feelings with me.
37. He is friendly and warm with me.
38. He just takes no notice of some things that I think or feel.
39. How much he likes or dislikes me is not altered by anything
that I tell him about myself.
40. At times I sense that he is not aware of what he is really
feeling with me.
41. I feel that he really values me.
42. He appreciates exactly how the things I experience feel to me.


4
1967). These variables, empathy, warmth, genuineness, and concrete
specificity appear to be integrally related to, and generally are
measured by, the degree of sensitive, appropriate responsiveness
demonstrated by the person in the counselor role.
The question which naturally emerged from the above was whether
pure communicationaI response style was sufficient to identify and pre
dict effective counseling. For example, if Porter's assumption that
understanding, interpretive, etc. responses facilitate beneficial
client change were empirically valid, then perhaps an investigation of
response preferences could reveal one crucial factor in effective coun
seling. In fact, it might be possible to provide stimuli similar to
the counseling situation and test on paper how a counselor candidate
or trainee would prefer to respond in some future counseling relation
ship. This study explored whether those assumptions were warranted in
reference to one particular instrument for measuring response preferences.
Statement of the Problem
This study has investigated whether or not a measurable relation
ship exists between the effectiveness ratings of a counselor in training
and certain interpersonaI communicationaI modes. The question was
examined utilizing two discrete criteria. One criterion related
preferred verbal responses to client perceptions of the relationship
shared with the counselor trainee. The second criterion explored the
relationship of counselor trainee response preferences to the evaluation
of effectiveness made by the counselor trainee's individual supervisor.
tore specifically, the questions answered by this study are:
I. Does a counselor trainee's response preferences on a printed


31
students. Using the same instrument, Keltz (1966) found the interjudge
agreement on pooled ratings to range between .73 and .78 and concluded
that the instrument and the supervisor rating approach "constitute a
usable, realistic measure for assessing the counseling proficiency of
counselor-trainees" (p. 516). Likewise, when Falkenberry compared the
evaluation consistency of supervisor, peer, role-player, and self ratings
on the Counselor Evaluation Form, he concluded that for overall scores
and in specific items the supervisors were more reliable than any of the
other groups of raters.
A comprehensive review of published studies using the Counselor
Evaluation Rating Scale is presented in the next chapter. The CERS
(Myrick & Kelly, 1971) is one of the criterion measures employed in
this study and, therefore, is explored in depth under the section on
"Instrumentation."
Cl ient Ratings
The use of cIient perceptions to evaluate counseling effectiveness
has been extensive, and the reactions of researchers overall tend to
qualify the approach as a valid criterion. The value of client ratings
is attested to by a substantial number of authors (Arbuckle, 1956; Form,
1955; Grigg, 1961; Grigg & Goodstein, 1957, 1959; Linden, Stone, &
Shertzer, 1965; Mueller, Gatsch, & Ralston, 1963; Patterson, 1958;
Pohlman, 1961; Pohlman & Robinson, I960; Rosen, 1967; Severinsen, 1966;
Shoben, 1953; Stablein, 1962; Thompson & Miller, 1970).
An early cogent argument for the inclusion of client ratings in
effectiveness research was penned by Grigg and Goodstein (1957). Their
position was that:


125
Stone, A. R., Frank, J. D., Nash, E, H., & Imber, S, D, An intensive
five-year follow-up study of treated psychiatric patients. JournaI
of Nervous and Mental Disorders, 1961, 155, 410-421,
Stoughton, R. W. The preparation of counselors and personnel workers.
Review of Educational Research, 1957, 27, 174-185.
Stripling, R. 0., & Lister, J. L. Selection, preparation and profession
alization of a specialist. Review of Educational Research, 1963,
33. 171-178.
Sullivan, H. S. The Interpersonal Theory of Psychiatry. New York:
W. W. Norton, 1953.
Sunderlund, D, M., & Barker, E. N, The orientation of psychotherapists.
Journal of Consulting Psychology, 1962, 26, 403-409.
Thompson, A., & Miller, A. Steps toward outcome criteria in counseling
and psychotherapy, 1970, ERIC ED 070 010.
Thoresen, C. E. Relevance and research in counseling. Review of Educa
tional Research, 1969, 39, 264-282. (a)
Thoresen, C. E. The systems approach and counselor education: Basis
features and implications. Counselor Education and Supervision,
1969, 9, 17-31. (b)
Thornton, B. M. Dimensions of perceived relationship as related to
marital adjustment. Unpublished master's thesis, Auburn Univer
sity, I960.
Tomlinson, T. M., & Hart, J. F., Jr. A validation study of the process
scale. Journal of Consulting Psychology, 1962, 26, 74-78.
Tos, D. J., Frumkin, R. M., & Wilson, M. E. Intercorrelations of four
relationship components of the Barrett-Lennard Relationship Inven
tory. Psychological Reports, 1968, 23, 641-642.
Truax, C. B. An approach toward training for the aide-therapist:
Research and implications. Address: Symposium on non-traditionaI
preparation for helping relationships, American Psychological
Association Convention, Chicago, September 5, 1965.
Truax, C. B. Comparisons between control patients, therapy patients
receiving high conditions, and therapy patients perceiving low
conditions on measures of constructive personality change. Brief
Research Reports, Wisconsin Psychiatric Institute, University of
Wisconsin, 1962. (a)
Truax, C. B. The process of group psychotherapy. Relationships between
hypothesized therapeutic conditions and intrapersonal exploration.
Psychological Monographs, 1961, 75. (7, Whole No. 511).


78
TABLE 13
REGRESSION EQUATION FOR THE COUNSELOR
EVALUATION RATING SCALE PREDICTORS
Scales Equations
Total Score

2.32 (Factor
E)
+ 140.10
Counseling Score
=
1,39 (Factor
E)
- 0.84 (Factor D)
+ 65.94
Supervision Score
(No Predictors)


84
reference, the communication style of the higher rated counselors
apparently was characterized by sensitivity and timing. The implica
tion is that the counselor cared enough to experience with and share
openly concerning the harsh, sometimes painful, realities of life.
Nevertheless, the helper transmitted, despite the uncomfortable con
tent, underlying caring for and commitment to the client.
It is interesting that Frank Feedback became a negative predictor
of the Unconditionality of Regard score. Perhaps this observation should
be interpreted in light of the growing disenchantment within counseling
theorist circles with the concept of unconditional regard. Some coun
selors question the possibility or even the appropriateness of communica
ting unconditional regard in the face of deteriorative or even destructive
client behavior. Certainly no definitive statement can be inferred from
this sparse evidence, but the field may continue to benefit from a
reexamination of the concept.
Unmistakably the most consistent predictor of BLRI scores was
graduate grade point average. This finding can be looked at in a number
of ways. In part the GPA had been influenced by practicum and intern
ship grades so the counselor trainee's GPA reflected some "expert"
evaluation of his or her counseling abiIity as we I I as a component of
cumulative experience. In part higher GPA suggests a higher verbal
facility which may be an indirect measure of the core concern of this
study, communication style. Another plausible explanation of the rela
tionship between high GPA and higher BLRI scores is a "prestige" factor.
Perhaps high GPA reflects an intelligence factor which impresses the
client. The counselor trainee with high GPA may be seen as one who is
knowledgeable, skilled, self-confident, and a worthwhile person with
whom to relate more closely.


2
been disappointing and inconclusive. One result has been that admission
criteria to counselor training has been severely criticized for showing
little relationship to effective counseling (Rapaport, 1968; Thoresen,
1969). These indictments appear warranted in light of the research which
suggests that counseling can indeed be "for better or for worse" (Car-
khuff, 1966; Lister, 1970). A mounting body of research indicates that
unless a counselor can embody personal fac Iitative qualities such as
empathy, warmth, genuineness, and concreteness, he may adversely affect
the client (Carkhuff & Berenson, 1967). The profession is challenged
by these findings to devise and employ more effective selection criteria
to increase the probability of a counselor making a beneficial rather
than a detrimental impact on the client (Thoresen, 1969).
Intellectual attributes have been used extensively to screen
candidates for counselor training programs. Unfortunately, while
ability measures may predict graduate grade point average with reason
able accuracy, neither aptitude test scores nor grade point averages
yield a significant correlation with counseling effectiveness (Car
khuff, Piaget, & Pierce, 1968; Bergin & Solomon, 1968). The Graduate
Record Examination, although widely used in graduate counselor educa
tion programs for selection purposes, appears to provide little meaning
ful predictive validity (Wittmer & Lister, 1971).
Since 1963, when the American Personnel and Guidance Association
encouraged a shift to the evaluation of personality variables in
counselor candidates, a vast array of personality correlates with
counseling effectiveness have been examined. Illustrative of this
attention to the personality of the counselor were the reviews of
Burnett (1954),Stoughton (1957), Hill and Green (I960), Stripling and


77
TABLE 12
REGRESSION EQUATION FOR THE BARRETT-LENNARD
RELATIONSHIP INVENTORY PREDICTORS
Dimensions
Equations
Tota 1 Score
80.
Level of Regard
=
2.
0.
44.
Empathy Level
=
19,
Unconditionality
of Regard
' 25.
Congruence
=
17.
10 (GPA) 3.62 (Factor F) 154.76
48 (Factor J) + 13.69 (GPA) +
53 (Factor A) 0.98 (Factor I) -
75
08 (GPA) 1.12 (Factor F) 32.44
13 (GPA) 2.01 (Factor J) 56.58
35 (GPA) 31.91


24
and Carkhuff is based upon these scales. Carkhuff, after the two
researchers went divergent paths, modified each of Truax's therapeutic
conditions scales to incorporate five levels for each scale (Carkhuff,
1969a; Carkhuff & Berenson, 1967). The five levels each represent a
qualitative degree of therapeutic functioning, allowing quantification
of the extent to which a therapist corresponds with the standardized
criteria. At first the scales were used to rate excerpts of audio
taped counseling interviews. More recently video tapes have been the
medium under investigation (Kagan & Krathwohl, 1967), particularly in
the evaluation of the InterpersonaI Process Recall (I PR) technique.
Even though the reliability and validity of the Carkhuff scales are
at times questioned (Hefele & Hurst, 1972), the summarization of the
extensive body of research (Carkhuff & Berenson, 1967) presents one of
the most persuasively supported theoretical positions in counseling
today. Carkhuffs scales are probably the most widely used rating
scales in current studies of self-exploration and the facilitative con
ditions of empathy, warmth, genuineness, and concreteness.
A central concept in the model for effective counseling which is
presently being explicated by Carkhuff is that facilitative conditions
can not only be measured but can also be transmitted to counselor
trainees by counselor educators who themselves provide them (Carkhuff &
Truax, 1966). Since the accumulating weight of research supports the
belief that counseling can be "for better or for worse," depending
upon the presence or absence of these dimensions, the implications of
this research for counselor selection and accountability monitoring are
clearly evident. Since the factor subscales of the predictor instrument
in this research relate directly to the antecedent research of Carkhuff


15
or "unobtrusive measures" with a variety of behaviors being studied.
Evaluation of counseling success has varied from dropout rate to atten
dance and grades to the proportion of time spent talking by the client.
A serious question, however, has been raised by Hill (I960) regarding
the validity of such outcome criteria. In reality such behavioral
criteria are responsive to change effects from a host of uncontrollable
and often unidentifiable variables.
In their I960 review of research, Hill and Green traced a large
portion of the difficulty in finding a consensual definition of effec
tiveness to the complexities of the settings within which counselors
function. They cited the Iiterature to support their contention that
the issues of selection and effectiveness are often confused. Large
variations in roles and relationships exist even in one counselors work
(APA, 1958; Hoffman, 1959), and across jobs the diversity of skills and
attitudes needed by counselors becomes staggering. There is growing
awareness that often, as Hill and Green (I960) assert, the counselor may
be called upon to fulfill responsibilities which the profession has
neither embraced nor prepared the counselor to do (Purcell, 1957;
Wellman, 1957). Today the role and function of the counselor appears to
be diverging even more from a precisely defined model which is amenable
to unambiguous evaluation.
Beyond the lack of agreement as to what is the preferred model of
effective treatment for a specific clientele, lies the diversity of
opinion about the different schools of therapy. In one study Sunderland
and Barker (1962) found that experienced therapists from Freudian,
Sullivanism, and Rogerian schools exhibited different behaviors, but
precisely how the specific counseling approach affects the effectiveness


55
been involved during the fall quarter. Data analysis yielded a .94
product moment reliability coefficient.
Myrick and Kelly (1971), in their presentation of the instrument,
reported they had failed to find relationships between either Graduate
Record Examination scores or undergraduate grade point averages and CERS
ratings. The correlation coefficients were respectively -.03 and -.25.
Wittmer and Lister (1971) also correlated Graduate Record Examina
tion scores and CERS ratings for 53 counselors in training whose CERS
ratings were compiled following the first quarter of practicum. The
GRE had been taken prior to admission. The correlation of the GRE
verbal score and CERS ratings was -.08. When the GRE quantitative score
was correlated with the CERS, the coefficient yielded was .12. Neither
correlation was significant at the .05 level. However, this same study
also compared results of each counselor trainees 16 PF profile, taken
the quarter before the first practicum, with his CERS ratings. The
authors employed the index of predicted counseling effectiveness which
is a regression equation based upon studies of the 16 PF at the Insti
tute for Personality and Ability Testing (1963). The Pearson product-
moment correlation coefficient produced was .41, which was significant
beyond the .01 level.
In another published article using the CERS as a criterion of
counseling effectiveness, Myrick, Kelly, and Wittmer (1972) found that
scores correlated significantly with certain factors of the 16 PF. The
investigators defined 20 effective counselor trainees with high CERS
scores and 20 ineffective counselor trainees with low CERS scores. The
15 trainees with scores near the mean of the CERS were eliminated. An
examination of the 16 PF scores of the two comparison groups yielded


Bailey, M. A., Warshaw, L., & Eichler, R. M. A study of factors related
to length of stay In psychotherapy. Journal of Clinical Psychology,
1959, J_5, 442-444.
Barahal, D. G., Brammer, L. M., & Shostrom, E. L. A client-centered
approach to educationaI-vocationaI counseling. Journal of Con
sulting Psychology, 1950, J_4, 256-260.
Barry, R., & Wolf, B. Five years of the Personnel and Guidance Journal.
Personnel and Guidance Journal, 1958, 36, 549-556.
Barrett-Lennard, G. T. Dimensions of therapist response as causal
factors in therapeutic change. Psychological Monographs, 1962,
76, No. 43.
Barrett-Lennard, G. T. Technical note on the 64-item revision of the
Relationship Inventory. Unpublished mimeograph report, Univer
sity of Waterloo, 1969.
Barrington, B. L. Prediction from counselor behavior of client percep
tion and of case outcome. Journal of Counseling Psychology, 1961,
8, 37-42.
Bartlett, W. E., & Thompson, C. L. Counselor preparation: A semantic
differential evaluation. Counselor Education and Supervision,
1971, JJ_, 129-133.
Bateson, G., Jackson, D. 0., Haley, J., & Weakland, J. Toward a
theory of schizophrenia. Behavioral Science, 1956, J_, 251-264.
Benton, J. A. Perceptual Characteristics of Episcopal Pastors.
Doctoral dissertation, University of Florida, 1964.
Berenson, B. G., & Carkhuff, R. R. The Sources of Gain in Counseling
and Psychotherapy. New York: Holt, Rinehart and Winston, Inc.,
1967.
Berenson, B. G., Carkhuff, R. R., & Myrus, P. The interpersonal func
tioning and training of college students. Journal of Counseling
Psychology, 1966, JJ5, 441-446.
Bergin, A. E., & Solomon, S. Personality and performance correlates
of graduate training. American Psychologist, 1963, _l_8, 393.
Berzon, B. The self-directed therapeutic group: An evaluative study.
LaJolla, California: Western Behavioral Sciences Institute
Reports, 1964, No. I
Bishop, J. Another look at counselor, client, and supervisor ratings
of counselor effectiveness. Counselor Education and Supervision,
1971, _[0, 319-322.


88
Research into the relationship of the HRPI to other instruments
used to evaluate counseling effectiveness as outlined in Chapter II may
be profitable. Specifically it may be productive to relate response
preferences to the Rokeach Dogmatism Scale, for example, or to any of
the personality measures such as the 16 Personality Factor Questionnaire
or the Edwards Personal Preference Schedule. It is conceivable that a
multiple criterion approach to counseling effectiveness might yield a
combination of personality variables plus communicationaI modes which
would produce a predictive formula robust enough to survive replication.
Finally, this present research study should be replicated. Repli
cations should be carried out, first of all, under approximately equiva
lent conditions to examine the reiabiIity of these findings. Next it
would be useful to conduct similar investigations with differing samples
such as paraprofessionaIs and career helper trainees from various dis
ciplines, representing a broad spectrum of theoretical orientations.
Perhaps it would be enlightening to carry out this research design using
the same predictor instrument but substituting different client and
supervisor ratings as criterion instruments. Clearly, this study and
parallel studies have hardly scratched the surface of what may prove
to be a fertile field of investigation.
Conclusions of the Investigation
With due regard for the possible limitations of the study cited
in Chapter III and expanded in this present chapter, there are certain
tentative conclusions which can be drawn. First, the Helper Response
Preference Inventory does appear to predict client perceptions of the
counseling relationship on the BLRI in the case of four factors which


41
is strictly for research purposes the ratings would be analyzed as group
data only.
In order to identify missing data, each Relationship Inventory
carried the name of the counselor trainee being evaluated. Five days
following the period of data collection, the researcher began calling
counselor trainees to identify and rectify the causes of unreturned
Relationship Inventories. Postcards were mailed to counselor trainees
who could not be reached at home to inform them of the missing Inventory
and request followup. Thirty days after the original week of data
collection it was assumed that all completed Inventories had been
received. Statistical analysis of the available data then proceeded.
Whenever counselor trainees were absent from their group supervi
sion sessions, they were contacted by telephone to solicit their coopera
tion. When the HRPI was administered and the Relationship Inventory
delivered to these trainees, the same instructions were given as in the
group sessions. Also, the follow-up of missing data was the same as
previously described for trainees instructed individually.
Limitations
Since a deliberate sample of counselor trainees was selected for
study rather than a random sample from the entire universe of counselor
trainees, the genera I izabiIity is limited to graduate counselor educa
tion students at the University of Florida, their supervisors, and their
clients. Furthermore, the exclusion of elementary and middle school
counselors, although methodologically essential, further minimized the
external validity of the results obtained.
A second limitation of this study lies in the level of development


Table 8Continued
HRPI E
HRPI F
HRPI G
HRPI H
HRPI 1
HRPI J
BLRI
Tota 1
0.05
-0.20
-0.01
-0.20
-0.03
0. 1 9
BLRI
Regard
0. 12
-0. 1 1
-0.01
-0.24
-0.06
0.41
BLRI
Empathy
0.03
-0.21
-0. 1 1
-0.24
-0.06
0.17
BLRI
Unconditiona1Ity
-0.01
-0. 1 1
0. 17
-0.03
-0.05
-0.13
BLRI
Congruence
0.09
-0. 17
-0.07
-0.16
0.06
0.25
CERS
Tota 1
0.27
0.04
0.01
-0.05
0.08
0.13
CERS
Counsel 1ng
0.32
0.00
-0.01
-0. 1 1
0.05
0. 10
CERS
Superv1sion
0. 18
0.08
0.00
0.02
0.10
0. 14
HRP 1
A
0.50
0. 1 1
0. 19
-0.39
0.08
-0.08
HRP 1
B
0.26
0.17
0.12
-0.16
0.41
0. 16
HRP 1
C
0.20
-0.00
0.50
-0.03
0.14
0.03
HRP 1
D
-0.02
0. 17
-0.14
0.05
-0. 19
0.14
HRP 1
E
1.00
0. 13
0.25
-0.20
0. 1 1
0.08
HRP 1
F
1 .00
-0.09
-0.06
0.22
-0.09
HRP 1
G
1.00
-0.04
0.01
0.06
HRP 1
H
1 .00
0. 10
0.00
HRPI
1
1.00
0. 16
HRPI J 1.00
as
\o


71
TABLE 9
MEAN CORRELATION OF
WITH ALL OF THE
TWO CRITERION
EACH HRPI FACTOR
SCORES OF THE
INSTRUMENTS
HRPI Factor
Mean of Correlation with All Scores
BLR I CERS
Factor
A
(Empathic Understanding -
Probing)
.16
.07
Factor
B
(Confrontation -
False Reassurance)
.01
.17
Factor
C
(Specificity Ambiguity)
.01
.21
Factor
D
(Analytical Interest -
Validation)
.05
-. 13
Factor
E
(Clarification -
Superficiality)
.06
.26
Factor
F
(Focused Probing -
Premature Judgments)
-.16
.04
Factor
G
(Tentative Formulations -
Avoidance of Emotions)
-.01
.00
Factor
H
(Requesting Giving
1nformation)
-.18
-.04
Factor
1
(Internal External Frame
of Reference)
-.03
.08
Factor
J
(Frank Feedback -
1ncongruence)
.18
.12
on the order of -.18 and .18 respectively merit consideration despite
the lack of precision afforded by this crude statistical treatment.


42
of the instruments employed as criteria of counseling effectiveness.
Neither the Barrett-Lennard Relationship Inventory nor the Counselor
Evaluation Rating Scale was validated directly upon client outcome
research. Therefore, the measurement of counseling effectiveness was
indirect, but, nevertheless, consistent with procedures generally
recognized as appropriate within the professional field of counseling.
One should not, however, construe the findings of this study to be
equivalent to validation based upon actual outcome measures of client
change. Likewise, the results of this study are specific to the instru
ments used and should not be generalized to other comparable instruments
until further study demonstrates which global measures of counseling
effectiveness may be considered statistically interchangeable.
Instrumentation
The three instruments referred to in the presentation of the metho
dology are each examined independently in the following section.
Helper Response Preference Inventory (HRPI)
The Helper Response Preference Inventory (HRPI) was originally
devised by Dr. James L. Lister in the early 1970's in an extension of
the work of Porter (1950) on classification of counseling responses.
The HRPI (see Appendix) consists of fifty items, each presenting a
statement by a he I pee to which the respondent must choose one of two
optional helper statements as a response. The intent of the forced-
choice format was to pair twenty responses, of approximately equal
valence, from each of the five following categories: understanding,
supportive, interpretive, probing, and evaluative responses. Thus, the
one hundred total responses enabled each set of possible pair-choices


3
Lister (1963), Froehlich (1951), Cash and Munger (1966), and Whiteley
(1969). Again, the results have been conflictual and inconclusive.
In his review of research in this area, C. H. Patterson concluded that
the data were sporadic, often poorly integrated with the basic issue
questions, and generally inadequate (1967).
An entirely different area of investigation began in 1950 with the
publication of the book Therapeutic Counseling by E. H. Porter. His
book presented the thesis that the counselors response pattern tends
to enhance or impede the flow of helpful communication. He contended
that the responses a counselor selects from his broad range of options
have implications for his ability to be an effective helping person.
His assertion seemed to imply that the primary index of a counselor's
ability to interact helpfully was the skill with which he could communi
cate with another human being. This appears to hold today.
In a similar vein, an entire school of counseling and psycho
therapy has developed around communication as the central concept in
human interaction (Bateson et al., 1956; Jackson, 1965; Haley, 1963;
Satir, 1967). While communication involves much more than simple
verbal exchanges, the communication theorists emphasize the importance
of the "report" or "message" component. Their belief is that largely
through verbal responses to himself, or "reflected appraisals" as
Sullivan (1953) called them, a person constructs his self concept, his
conception of the outside world including percepts of other persons,
and his sense of identity based upon how he personally fits into that
outside world. These theoretical constructs have been supported in
part by the process research on counseling which has linked at least
four variables to beneficial client outcome (Carkhuff & Berenson,