Review of Research

on Practicum Supervision
JAMES c. HANSEN
RONALD POUND
CAROLE PETRO

This five-year review of research on practicum supervision examines research

findings and research designs. Research is classified according to the role of the
supervisor, the training procedure, and the rating systems. The major focus of
research has been on training in the core conditions of facilitative communi-
cation. Modeling techniques have been used effectively in practicum, didactic
approaches have been more effective than experiential approaches, and audio-
tape procedures in training were as effective as videotape techniques.

This five-year review assesses the findings of practicum supervision re-

search published from 1970 through 1974; it follows the ten-year review by
Hansen and Warner (1971). The current survey indicates a substantial
increase in the volume of research articles published. There were 25 articles
reported in the earlier review; 29 have been reported during the past five
years. All APA- and APGA-affiliated journals were examined; research
articles were found only in Counselor Education and Supervision (19)and
the Journal of Counseling Psychology (11). Only articles reporting
research projects related to practicum supervision were reviewed.
The articles are grouped into topical divisions determined inductively
from the major focus of the study. Articles classified under "Role of the
Supervision" include those focusing on the supervisory relationship and
supervisory behavior. "Training Procedures in Supervision" emphasizes
the use of video and audio techniques and training in communication. The
'Rating Systems in Practicum" category include articles on techniques to
evaluate counselors. -

ROLE OF THE SUPERVISOR

Eleven articles concentrated on some aspect of the role of the supervisor. An
experimental design was employed in seven studies, thereby increasing the
generalizability of the findings. All but one study explore the relationship

James C. Hansen is Professor of Counselor Education and Carole Petro is Assistant Dean of
the Colleges, both at the State University of New York-at Buffalo, Amherst, New York. Ronald
Pound is Assistant Professor, Southern Methodist University, Dallas, Texas.

DECEMBER 1976 107

between supervisor behaviors and the trainee's learning of facilitative com-
munication.
Payne, Winter, and Bell (1972) found that trainees in the technique-
oriented supervision and two control groups gave higher levels of accurate
empathy to recorded client statements than did trainees in the counseling-
oriented supervision. The presupervision use of audio modeling of empa-
thic responses also increased the trainee's facilitative level of empathy. In a
similar study, Payne, Weiss, and Kapp (1972) reported that audio modeling
of empathy and didactic supervision were significantly related to the learn-
ing of empathic responses by trainees. Moreover, nonsupervised controls,
in either the modeling or no-modeling conditions, did not change signifi-
cantly in their ability to offer empathy to taped client statements.
Does matching the method of supervision with supervisee preference
in supervision increase the level of empathic responses? According to Birk
(1972), matching of supervisee preference to supervisory method did not
increase the supervisee's level of empathy toward the client. The didactic
approach was significantly more effective than the experimental.
Lanning's (1971) comparison of individual and group supervision
revealed no significant difference in the client's perception of the trainee's
counseling relationships, as measured by the Barret-Lennard Relationship
Inventory. Knowing how the trainees perceived their supervisors explained
half the variance in how the trainees expected their client to perceive them.
Trainees expected a-relationship with the client that was similar to their
perceived relationship with the supervisor.
Sunblad and Feinberg (1972) randomly assigned supervisees to one of
three experimental groups and received a set of either positive, neutral, or
negative expectations about the supervisee. The favorableness of expecta-
tions held by one member of a dyadic relationship was considered an inter-
personal attraction. The results indicated that supervisor warmth is posi-
tively affected by interpersonal attraction, but supervisor empathy or
genuineness is not. Interestingly, a positive set of expectation combined
with experience in years yielded high process dimensions, while negative
expectations resulted in lower levels of process dimensions for experienced
supervisors than for their less experienced counterparts.
Lambert (1974) showed that, although genuineness and respect are
offered at similar levels in counseling and supervision, levels of empathy
and specificity are significantly lower in supervision than in counseling.
An analysis of the data on the Hill Interaction Matrix suggests that supervi-
sors frequently ignore the trainee's feelings and focus instead on the feel-
ings of the trainee's client.
According to Pierce and Schauble (1970), significant gains in
therapist-offered facilitative conditions were made by supervisees who had
high-level supervisors; supervisees with high-level supervisors functioned
significantly higher on the dimensions of empathy, regard, genuineness,

108 COUNSELOR EDUCATION AND SUPERVISION

and concreteness than did supervisees of low-level supervisors; and super-
visees of low-level supervisors gained only if they were functioning at a
lower level than their supervisor. Furthermore, the assessment of trainee
change on core dimensions was detectable early in the supervisory relation-
ship. In a follow-up study to determine the durability of effects, Pierce and
Schauble (197la) found no change in the facilitative levels of supervisees
who had received high-level conditions in supervision. Counselors who
were functioning at lower levels of facilitation after supervision with low-
level supervisors similarly showed no differences from ratings in the origi-
nal study, except for positive gains in concreteness. Another study by Pierce
and Schauble (1971 b) generally supported their earlier findings. Students
with high-level practicum instructors and high-level individual supervi-
sors showed significant change in their capacity to offer the facilitative con-
ditions. However, students with high-level practicum instructors and low-
level individual supervisors showed no significant change in their level of
facilitation.
Austin and Altekruse (1972) indicated that trainees in a group practi-
cum with a nonrestrictive supervisor emitted significantly more under-
standing responses at the end of the practicum than trainees with a restric-
tive group practicum supervisor. Nonrestrictive supervisors were identi-
fied as leader centered. No differences were found in the trainee's ability to
emit affective and exploratory responses in their final counseling inter-
views. Interestingly, the highest percentage of understanding responses
were obtained by the control group.
Comparing monitor-modeling and immediate feedback supervision,
Silverman and Quinn (1974) reported that monitor-modeling supervision,
in which the supervisor is present during the interview, significantly
increased the level of facilitative conditions offered by the trainee to a
coached client. The immediate feedback treatment, involving supervisor-
initiated tape analysis and direct suggestion, accounted for no significant
changes in the pre and posttest measures of facilitation. The study suggests
that collaborative or co-counselor techniques are less threatening to the
trainee.
Conclusions
A disproportionate amount of research effort is given to facilitative com-
munication. Ten of the eleven studies concerned with the role of the super-
visor examined aspects of communicating core conditions. Other aspects of
the supervisor's role and other criteria need examination. The results indi-
cated that the trainees' level of facilitative communication can be improved
through a variety of supervisory experiences. Two interesting general con-
clusions are (a) that the supervisor who directly or indirectly models facilit-
ative behavior will be more effective and (b) that didactic training is more
effective than experiential training.

DECEMBER 1976 109

TRAINING PROCEDURES IN SUPERVISION

Fourteen articles focused on the effects of training techniques or counselor

techniques on counselor behaviors. They are designated as training stud-
ies because they emphasized the impact of techniques in supervision rather
than the supervisors' behaviors. Some studies examined training tech-
niques within an actual practicum experience; others investigated similar
training procedures that have potential for application in the practicum.
Nine studies used videotape techniques, one used galvanic skin response
(GSR) equipment, and four dealt with short verbal training programs.
Experimental designs were used in nine of the fourteen investigations.
Yenawine and Arbuckie (1971) have maintained that feedback to train-
ees using videotape is both appropriate and meaningful in practicum
supervision. However, early in the training period, the use of audio rather
than video technique is optimal. Their conclusions were based on subjec-
tive log reports of students in either an audio- or video-oriented practicum.
Ward, Kagen, and Krathwohl (1972) found no significant differences in
counseling effectiveness among practicum students assigned to either a
video interpersonal process recall group, an audio interpersonal process
recall group, or traditional supervision (i.e., using an audiotape). The
results of the investigation of the relative effectiveness of different methods
of playback on student counselor performance indicated no significant dif-
ferences among the audio/video, audio, video, and control (no playback)
groups (Markey et al. 1970).
Saltmarsh's (1973) data supported the use of a two-media programmed
instruction to increase levels of empathy in counseling activities. The con-
trol group, which participated in discussions based on readings and viewed
a videotape, scored significantly lower in their empathic responses to
videotaped counseling sessions than did the experimental group.
Spivak (1972) reported that the achievement of interviewing skills by
trainees in an experimental prepracticum was significantly greater than
the achievement of trainees in a traditional prepracticum. The experimen-
tal prepracticum was based on counselor developmental tasks and imple-
mented by the interpersonal process recall technique; the traditional pre-
practicum consisted of an overview of counseling theories and practices
that made extensive use of group discussion and critiques of audio and
video tapes.
Using client perceptions of relationship variables as the criterion,
Bradley's (1974) study revealed no significant differences among the core
dimensions offered by trainees exposed to a modified form of interpersonal
process recall and controls in a traditional practicum setting.
Boyd (1973) showed that variance of the participants' scores on the
Counselor Verbal Response Scale was not significant between two micro-
model treatments using either behavioral techniques or recall interroga-

110 COUNSELOR EDUCATION AND SUPERVISION

tion. Significant differences were found between micromodel participants
and controls. Behavioral supervision was associated with greater trainee
gains on two of four measures on the Counselor Verbal Response Scale.
In an experiment based on principles of social learning, Dalton,
Sundblad, and Hylbert (1973) demonstrated that the participants in a
modeled learning experience communicated empathic understanding at
significantly higher levels than either the treatment controls who partici-
pated only in part of the modeled experience or the regular controls. A one-
month follow-up study indicated no significant loss of functioning.
Butler and Hansen (1973) reported that levels of facilitation can be
increased significantly in a brief didactic/experiential training program,
when facilitation is assessed from a coached client interview or by written
response to taped client statements. Skills acquired by the experimental
group were retained during a four-week latency period, which implys that
the skills were integrated into counselor response patterns. Lack of train-
ing in the control group led to no significant differences in the functional
level among prerated, moderate, and low-level counselors.
In an examination of the relative effectiveness of immediate verbal
feedback, and feedback and instructions on the qualitative levels of empa-
thic verbal performance of counselor trainees, Carlson (1974) showed that
the most effective way of increasing empathic responses was to present the
counselor trainee with immediate feedback in combination with instruc-
tions. All treatment effects significantly increased the trainees' responses
when compared to counselors in the control group receiving the usual
delayed feedback procedures.
Guttman (1973) found that participants trained in nondefensive com-
munication demonstrated a significant reduction in defensive behavior
during supervision when compared to participants who did not receive
nondefensive communication training.
Carkhuff, Fiel, and Kratochvil (1970) -investigated the differential
effects of sequences of training in counselor-responsive and counselor-
initiated dimensions. Both sequences in the counterbalance design demon-
strated significant pre to posttest changes in communication on the
counselor-responsive and counselor-initiated dimensions. Although the
counselor-responsive/counselor-initiated (R-I) sequence resulted in sig-
nificantly greater changes in the communication and discrimination of
counselor-responsive dimensions, the R-I sequence did not produce
similar results for the counselor-initiated dimensions. They concluded that
the greatest changes can be effected in the briefest training time by focusing
on counselor responsiveness before counselor initiation.
Conclusions
In general, videotape techniques are valuable aids in counselor training.
However, in most instances these studies revealed no significant differences

DECEMBER 1976 III

between audio and video techniques on trainee response criteria. As in the
first section, the effectiveness of modeling was supported by articles
reviewed in this category. Specific training programs are evidently success-
ful in teaching communication skills. The heaviest concentration of
research interest lay in facilitative communication, with seven of the four-
teen studies selecting some aspect of facilitative communication as a criter-
IOn.

RATING SYSTEMS IN PRACTICUM

Four articles, all of which utilized pre-experimental designs, were con-

cerned with rating systems in counselor training. Jansen, Robb, and Bonk
(1972) have suggested that three distinct patterns of characteristics typified
"A," "B," and "C" male practicum students. "A" students were more
objective, less hypersensitive, and more understanding of people than were
"B" and "C" students. Though not significantly different in emotional sta-
bility, B students were less objective and more hypersensitive than the A
students. The C students were described as "mostly touchy, hypersensitive,
and lacking in objectivity." Significantly greater scholastic aptitude and
significantly better cumulative grade point averages were characteristic of
A males in comparison with Band C males. They concluded that grades
generally are valid estimates of "potential for candidates in the counseling
profession" (p. 29).
Engle and Betz (1971) studied differences in peer and supervisory rat-
ings of counselor trainees in a regular practicum program and in a guid-
ance institute. In both groups, the correlation of peer and supervisory rank-
ings were highly significant. The test for differences in the magnitude of
the two rank correlations indicated that judgments of peers and supervisors
in the institute were correlated significantly higher than judgments of
peers and supervisors in the regular program.
Using Ivey's Rating Scale of Counselor Effectiveness as the criterion
measure, Friesen and Dunning (1973) found that interrater reliability was
high within three groups: practicum students (peers), lay people, and pro-
fessional counselors and supervisors. In addition, there was strong agree-
ment among groups on rank order of counselor effectiveness in the video-
taped interviews. A comparison of actual rather than rank-order scores
reflected significantly different ratings by practicum students from super-
visors and the lay people. The students rated their peers consistently higher
than they rated the other groups. Moreover, agreement between the lay and
supervisory groups suggests that formal training does not affect the ability
to assess counselor effectiveness as measured on Ivey's scale.
Bishop (1971) compared the perceptions of counselors, clients, and
supervisors with regard to the effectiveness of counseling as measured by
the Counseling Evaluation Inventory (CEI). Their results indicated that

112 COUNSELOR EDUCATION AND SUPERVISION

the counselor self-ratings and the supervisors' ratings significantly corre-
lated with each other, but neither significantly correlated with the client
CEI ratings. In addition, Bishop noted that the client CEI ratings were sig-
nificantly higher than both the counselor self-ratings and the supervisors'
ratings.

Conclusions
One study in this section found that various personality characteristics of
trainees are related to practicum grades. Similar to the findings of earlier
research, peer rankings of trainees tend to be highly correlated, and peer
rankings and supervisor rankings agree.

COMMENT

Four major points have been substantiated by these research projects: (a)
Modeling techniques can be used effectively in the practicum: (b) when
didactic and experiential approaches to training were compared, didactic
approaches were found to be more effective; (c) audiotape procedures were
as effective as videotape techniques; and (d) training in the core conditions
of facilitative communication has been the major concentration of supervi-
sion research.
Using the method of design evaluation developed by Campbell and
Stanley (1963), it appears that the sources of invalidity in supervision
research have been reduced by more rigorous designs, compared with
supervision research reviewed five years ago. Of the 29 studies examined
here, 19 employed an experimental design-either a posttest-only control
group design or a pretest-posttest control group design. One study used a
quasi-experimental, counterbalanced design; the remaining nine studies
were pre-experimental.
In spite of the upgrading of methodology, serious questions about the
conclusions from current research persist. In several instances, the lack of
an adequate sample size and the lack of specificity in variable definitions
limit the generalizability of findings. The experimental situation is gener-
ally not sufficiently rigorous to control for the effects of variables other
than the treatment. Caution is warranted in the interpretation or applica-
tion of results from supervision research. Furthermore, a more complete
description of procedures to permit replication and improve application is
needed in further studies. Further investigations should employ several
groups for each treatment condition to allow for more potent statistical
analyses.
Of the 25 studies on process and training, 17 examined some aspect of
facilitative communication. The communication of empathy was usually
the dependent variable. Several approaches to supervision have demon-

DECEMBER 1976 113

strated success in teaching trainees to respond at high levels of core condi-
tions. Apparently, supervisors who function at high levels of core condi-
tions stimulate those response patterns in trainees. Counselor trainees can
learn facilitative responses through more didactic techniques, the use of
models, and video and audio feedback procedures.
Although teaching facilitative communication is necessary, it is not
sufficient; research on supervisory procedures that improve other counsel-
or skills is needed. As Gormally and Hill (1974) and Matarazzo (1971) have
suggested, the more important issue may be whether the effectiveness of
supervision should be assessed in the trainee or in those the trainee helps.
The most effective step for future research in supervision would be to
explore the direct relationship between process variables in supervision
and the trainee's effectiveness as a counselor. Ideally, these investigations
should measure not only counseling process but counseling outcomes for
the trainee.
The initial steps have been taken toward explaining the specifics of the
supervisory process. The appearance of several studies using behavioral
techniques (e.g., modeling) signals the potential infusion of behavioral
methods into supervision research and practice. Undoubtedly, the nature
of the outcome and process of counselor training can be explored best by
incorporating behavior research practices and techniques. The result of
this approach is the systematic investigation of the complex processes
found in supervisory relationships. When systematic investigations begin
to characterize the literature, research findings will lead to a more efficient
use of counselor training techniques.
More planned, comprehensive research is preferable to isolated stud-
ies. Research serves as a link between past findings and current thinking;
new hypotheses are a by-product of earlier results. Investigations of this
nature become a part of the thinking and experimentation that follows and
thus are a vital link in the delineating process. Current research that is not
based on a thorough review of past works or that fails to replicate the exper-
imental conditions, becomes an isolated entity, at best accidentally rele-
vant. This type of research has created haphazard and unreliable growth in
the field of supervision.
REFERENCES
Austin, B., & Altekruse, M. K. The effect of group supervision roles on practicum students'
interview behavior. Counselor Education and Supervision, 1972,12(1),63-68.
Birk, J. M. Effects of counseling supervision method and preference on empathic understand-
ing. Journal of Counseling Psychology, 1972,19(6),542-546.
Bishop, J. B. Another look of counselor, client, and supervisor ratings of counselor effective-
ness. Counselor Education and Supervision, 1971,10(3),319-323.
Boyd, J. D. Micro-counseling for a counseling-like verbal response set: Differential effects of
two micromodels and two methods of counseling supervision. Journal of Counseling
Psychology, 1973,20(1),97-98.
Bradley, F. O. A modified interpersonal process recall technique as a training model. Coun-
selor Education and Supervision, 1974,14(1), 34-39.

114 COUNSELOR EDUCATtoN AND SUPERVISION

Butler, E. R., & Hansen, J C. Facilitative training: Acquisition, retention, and modes of
assessment. Journal of Counse.ling Psychology, 1973,20(1),60-65.
Campbell, D. T., & Stanley, J. C. Experimental and quasi-experimental designs for research.
Chicago: Rand McNally, 1963.
Carkhuff, R. R.; Fiel, T.; & Kratochvil, D. The differential effects of sequence of training in
counselor-responsive and counselor-initiated conditions. Counselor Education and
Supervision, 1970,9(2), 106-109.
Carlson, K. W. Increasing verbal empathy as a function of feedback and instruction. Counsel-
or Education and Supervision, 1974,13(3),208-213.
Dalton, R. F.; Sundblad, L. M.; & Hylbert, K. W. An application of principles of social learn-
ing to training in communication of empathy. Journal of Counseling Psychology, 1973,
20(4),378-383.
Eisenberg, S., & Delaney, D. J. Using video simulation of counseling for training counselors.
Journal of Counseling Psychology, 1970,17(1), 15-19.
Engle, K. B., & Betz, R. L. Peer ratings revisited. Counselor Education and Supervision, 1971,
10(2), 165-170.
Friesen, D. D., & Dunning, G. B. Peer evaluation and practicum supervision. Counselor Edu-
cation and Supervision, 1973,12(3),229-234.
Garmally, J., & Hill, C. E. Guidelines for research on Carkhuff's training model. Journal of
Counseling Psychology, 1974,21(6),539-547.
Guttman, M. A. J. Reduction of the defensive behavior of counselor trainees during counsel-
ing supervision. Counselor Education and Supervision, 1973,12(4), 294-299.
Hansen, J C., & Warner, R. W. Review of research on practicum supervision. Counselor Edu-
cation and Supervision, 1971,10, 261-272.
Jansen, D. G.; Robb, G. P.; & Bonk, E. C. Differences among male A, B,andCmaster'sdegree
candidates in a counseling practicum. Counselor Education and Supervision, 1972,12(1),
24-30.
Lambert, M. J Supervisory and counseling process: A comparative study. Counselor Educa-
tion and Supervision, 1974,14(1),54-60.
Lanning, W. L. A study of the relation between group and individual counseling supervision
and three relationship measures. Journal of Counseling Psychology, 1971,18(5), 401-
406.
Markey, M. J; Frederickson, R. H.; Johnson, R. W.; & Julius, M. A. Influence of playback
techniques on counselor performance. Counselor Education and Supervision, 1970,9(3),
178-182.
Matarazzo, R. G. Research on the teaching and learning of psychotherapeutic skills. In A. Ber-
gin & S. Garfield (Eds.), Handbook of psychotherapy and behavior change. New York:
Wiley, 1971. Pp. 895-924.
Michels, T. J; Gazda, G. M.; & Wiggins, S. Instrumentation (GSR) and its effects on counsel-
or responses. Counselor Education and Supervision, 1971, 10(3), 303-309.
Payne, P. A.; Weiss, S. D.; & Kapp, R. A. Didactic, experiential, and modeling factors in the
learning of empathy. Journal of Counseling Psychology, 1972,19(5),425-429.
Payne, P. A.; Winter, D. E.; & Bell, G. E. Effects of supervisor style on the learning of empathy
in a supervision analogue. Counselor Education and Supervision, 1972,11(4),262-269.
Pierce, R. M., & Schauble, P. G. Graduate training of facilitative counselors: The effects of
individual supervision. Journal of Counseling Psychology, 1970,17(3),210-215.
Pierce, R. M., & Schauble, P. G. Follow-up study on the effects of individual supervision in
graduate school training, Journal of Counseling Psychology, 1971,18(2), 186-187. (a)
Pierce, R. M., & Schauble, P. G. Toward the development of facilitative counselors: The
effects of practicum instruction and individual supervision. Counselor Education and
Supervision, 1971,11(2),83-89. (b)
Saltmarsh, R. E. Development of empathic interview skills through programmed instruction.
Journal of Counseling Psychology, 1973,20(4),375-377.
Silverman, M. S., & Quinn, P. F. Co-counseling supervision in practicum. Counselor Educa-
tion and Supervision, 1974,13(4),256-260.
Spivak, J D. Laboratory to classroom: The practical application of IPR in a master's level
prepracticum counselor education program. Counselor Education and Supervision,
1972,12(1),3-16.
Sundblad, L. M., & Feinberg, L. B. The relationship of interpersonal attraction experience

DECEMBER 1976 115

 and supervisor's level of functioning in dyadic counseling supervision. Counselor Edu-
cation and Supervision, 1972,11(3), 187-193.
Ward, G. R.; Kagan, N.; & Krathwohl, D. R. An attempt to measure and facilitate counselor
effectiveness. Counselor Education and Supervision, 1972,11(4), 179-186.
Yenawine, G., & Arbuckle, D. S. Study of the use of videotape and audiotape as techniques in
counselor education. Journal oj Counseling Psychology, 1971,18(1), 1-6.

Corrections

Janice Hildreth, Assistant Head Resident, Idaho State

University, Pocatello, was co-author of" 'The Candidacy
Game': A Simulation Game for Counselor Trainees,"
which appeared in the June 1976 issue of the Journal.

The title of the article by Elsie J. Smith in the September

1976 issue of the Journal (Vol. 16, No.1, pp. 13-24) should
read, "Issues and Problems in the Group Supervision of
Beginning Group Counselors."

116 COUNSELOR EDUCATION AND SUPERVISION