The development and validation of the group supervision scale

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content THE DEVELOPMENT AND VALIDATION OF THE GROUP SUPERVISION
Copyright2002
SCALE
By
Ferdinand Arcinue
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNNERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements of the Degree
Doctor of Philosophy
EDUCATION
(COUNSELING PSYCHOLOGY)
December 2002
Ferdinand Arcinue
UNIVERSITY OF SOUTHERN CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES, CALIFORNIA 90089-1695
This dissertation, written by
under the direction of hJ2.._ dissertation committee, and
approved by all its members, has been presented to and
accepted by the Director of Graduate and Professional
Programs, in partial fulfillment of the requirements for the
degree of
DOCTOR OF PHILOSOPHY
Director
Date December 1 8, 2002
Chair
List of Tables
Abstract
Chapter I
Introduction
Literature Review
Table of Contents
Differences between Group and Individual Supervision
Studies in Group Supervision
Models of Group Supervision
The Role of the Group Supervisor
Strategies for Group Supervisors
Strategies Related to the Developmental Level of the Supervisee
Specific Skills of the Group Supervisor
Group Development Over Time
Purpose of this Study
Group Climate
Satisfaction with Supervision
Skill and Conceptualization Development
Age, Physical Attractiveness, and Professional Appearance
Summary and Hypotheses
Chapter II
Method
Study 1
Item Development
Expert Raters
Study 2
Measures
Demographic Questionnaire
Group Climate Questionnaire-Short Form
Supervision Satisfaction Questionnaire
Supervisory Working Alliance Inventory-Client Focus subscale
Procedure
Chapter III
Results
Factor Analysis
Preliminary analyses
Descriptive Statistics
Construct Validity
Group Supervision Scale and Group Cohesion
Page
iv
v
1
4
5
14
19
23
27
31
34
35
37
38
41
42
44
44
45
45
45
47
49
50
50
51
51
51
52
54
54
58
58
62
63
ii
Group Supervision Scale and Skill/Case Conceptualization Development 64
Group Supervision Scale and Satisfaction with Supervision 66
Group Supervision Scale and Attractiveness, Appearance, and Age 67
Interaction Effects 68
ChapterN
Discussion 70
Group Supervision Scale and Group Cohesion 74
Group Supervision Scale and Satisfaction with Supervision 76
Group Supervision Scale and Skill and Case Conceptualization 76
Development
Group Supervision Scale and Age, Attractiveness, and Appearance 78
Implications for Theory, Research, and Practice 79
Limitations of the Study 82
References 88
Appendices
Appendix A 95
Appendix B 105
iii
List of Tables
Table Page
I Factor Loadings, Eigenvalues, Variance Explained, and Item- 57
Scale Correlations in the Group Supervision Scale
2 Means, Standard Deviations, and Intercorrelations 60
3 Results of the Univariate Simultaneous Regression Analyses: 3 63
Factors of the GSS and Group Cohesion, Skill Development, and
Satisfaction with Supervision
4 Two Way ANOV A for Main Effects and Interactions 68
lV
Abstract
The pmpose of this study was to develop and provide preliminary validation
evidence for the Group Supervision Scale (GSS), a measure to assess a supervisee's
perception of the group supervisor. Items were developed using focus groups
consisting of graduate student supervisees at the predoctoral intern level and expert
supervisors. After the scale was developed, the investigator engaged in a multi-step
process to validate the scale. The first step used an exploratory factor analysis to
determine the number of dimensions or factors involved in the evaluation of a group
supervisor. The second step focused upon measuring how the performance of the
group supervisor related to other constructs associated with group supervision. In
particular, to provide evidence for convergent validity the constructs of supervisee's
perception of group climate, supervisee satisfaction with supervision, and the
emphasis the supervisors placed on promoting the trainee's understanding of the
client and development of counseling skills were analyzed with regard to their
relationship to the GSS. In addition, the GSS was compared to the supervisor's age,
level of physical attractiveness, and appearance professionally to determine the
discriminant validity of the instrument.
v
Chapter I
Introduction
In the first review of the group supervision literature Holloway and Johnston
(1985) titled their article, "Group supervision: widely practiced but poorly
understood." A decade later, in a second review of the group supervision literature
Prieto (1996) summarized, "Although the use of supervision seems to be widespread,
relatively little research has focused specifically on the group supervision of
trainees" (p. 295).
The use of group supervision is widespread in training professional counselors.
In their national survey of predoctoral psychology internship programs, Riva and
Cornish (1995) found that 65% of the sites conducted group supervision. Because of
group supervision's widespread use and the relatively small base of validating
research, Holloway and Johnston (1985) stated that the continuing lack of empirical
research investigating the group supervision of trainees learning individual
psychotherapy is "indefensible" (p. 305).
Group supervision has many unique benefits that are different than those found
in individual supervision. The group format is more economical in that several
trainees can be supervised by one person at the same time, and the group members
can learn from each other. In addition, researchers have pointed out other
advantages of group supervision which include peer feedback and support, exposure
to a wider variety of diagnostic and treatment issues, and opportunities for personal
1
insight into interpersonal behavior (Bernard & Goodyear, 1998; Hayes, Blackman, &
Brennan, 2001; Tebb, Manning, & Klaumann, 1996).
Group supervision also has potential disadvantages in comparison to individual
supervision (Counselman & Gumpert, 1993). For example, due to time constraints
members may not have a chance to present their cases every week or in sufficient
detail, or personality conflicts may arise and disrupt the functioning of the group. In
addition, Bernard and Goodyear (1998) point out that individual supervision is still
considered the cornerstone of professional development, and Blocher (1983)
observed that authors often make reference to some form of group supervision, but
that it is frequently viewed as an adjunct to individual supervision.
Regardless of the potential disadvantages, the general belief among clinicians
and researchers is that group supervision has a great deal to offer (Bernard &
Goodyear, 1998). However, given both Holloway and Johnston's (1984) and
Prieto's (1994) fmdings regarding the lack of empirical research supporting the use
of group supervision, and, paradoxically, the wide spread use of this format, it has
become more apparent that there is a need for research to be conducted in this area.
From their review of the literature, Holloway and Johnston (1985) concluded
that "it is clear from a review of literature that the field of counselor training is at a
very rudimentary level of explaining and understanding group supervision" (p. 338).
Holloway and Johnston (1985) asserted that it is important and necessary for
researcher to begin a more systematic examination of the group supervision process
2
and that "without a systematic effort, group supervision will remain a weak link in
our training programs, widely practiced and poorly justified" (p. 339).
In his later review of empirical studies in group supervision, Prieto (1996)
found that the majority of the studies employed the use of"author-devised"
questionnaires and often did not report the psychometric properties of the instrument.
Prieto (1996) noted that most studies of group supervision are hampered by serious
methodological problems such as low sample size, subjective evaluation or
construction of dependent measures, and questionable establishment of treatment
conditions.
Bernard and Goodyear (1998) noted that almost all instruments that have been
used to assess supervision variables originally were developed especially for
psychotherapy research. This practice may change the psychometric properties of
the instrument. It also perpetuates the use of roles and metaphors from interventions
other than supervision. It is important to note that, of the small number of
instruments that have been developed specifically for supervision research, only one
has been developed to address the unique dynamics involved in group supervision
(Calsyn, Burger, & Winter, 1999).
Prieto (1996) stated that investigators need to develop and use supervision
specific assessment instruments in future research on group supervision. Holmes et
al. (1998) conducted their study using qualitative techniques, but also concluded that
future research should focus on the development of assessment instruments that can
enable the analysis of the complex interactions that occur in group supervision. In
3
this review of the literature, a computerized search revealed one study involving the
development of an instrument specific to group supervision (Calsyn, Burger, &
Winter, 1999). These researchers used generalizability theory to evaluate the
psychometric properties of a scale that assessed content and supportiveness of both
the leader and peers in group supervision. This instrument consisted of only four
items, and had limited utility.
The overall purpose of the study was to develop and provide preliminary
validation evidence of the Group Supervision Scale (GSS). The GSS is a measure
that evaluates the group supervisor on three factors: skill and case conceptualization
development, the facilitation of group safety, and group management (e.g., direction,
structure, time management).
Literature Review
In their review of the literature, Holloway and Johnston (1985) defined group
supervision as a process '!in which supervisors oversee a trainee's professional
development in a group of peers" (p. 333). For the purposes of this study, Bernard
and Goodyear's (1998) more comprehensive definition of group supervision will be
used:
Group supervision is the regular meeting of a group of supervisees
with a designated supervisor, for the purpose of furthering their
understanding of themselves as clinicians, of the clients with
whom they work, and/or of service delivery in general, and who
are aided in this endeavor by their interaction with each other in
the context of group process (Bernard & Goodyear, 1998, p. 111 ).
4
Differences between group and individual supervision
Riva and Cornish (1995) outlined the unique benefits of group supervision that
had been described in the literature. These include peer feedback and support,
exposure to a wider variety of diagnostic and treatment issues, and opportunities for
personal insight into interpersonal behavior. Tebb et al. (1996) noted that group
supervision provides a medium for forms of "mutual aid:" sharing data, presenting
ideas and having the ideas challenged, discussing counter-transference issues,
normalizing each others experiences, providing support and praise, and receiving and
giving feedback.
In their national survey of predoctoral psychology internship programs, Riva
and Cornish (1995) asked respondents an open ended question regarding the unique
contributions of group supervision in comparison to individual supervision. The
three most cited contributions were "provides diverse viewpoints and multiple
hypotheses," "provides vicarious learning," and "allows for group process issues" (p.
524). Although efficiency is sometimes thought to be a primary advantage of group
over individual supervision, these authors noted that few respondents cited this
variable.
Unlike individual supervision, groups provide for the possibility of direct peer
support. Westwood (1989) stated that the group is a particularly important source of
support for fledgling counselors, especially. Borders (1991) stated, "Peer groups
provide a supportive environment and reassurance that others are experiencing
similar feelings and concerns" (p. 248). Starling and Baker (2000) found that
5
supervisees emphasized the importance of peer feedback on their skill development
and self-confidence, and concluded that peer feedback and support are very
important ingredients of group supervision. In a similar fashion, Borders (1991)
stated that honest and constructive feedback from the other group members is crucial
to the success of the supervision group.
Hayes et al. (2001) noted that one of the important outcomes of participation in
a group is the opportunity for members to test their perceptions and improve their
communications with others. Group supervision provides supervisees the
opportunity to receive validation for their own ideas in the company of their peers.
According to Borders (1991), group supervision may be the only opportunity that
many counselors have for feedback on their effectiveness. When a supervisee
presents a case or intervention well, praise from several people has more impact than
from just one supervisor. The group allows for the supervisee to receive
reinforcement and feedback from a variety of sources (Westwood, 1989).
Walter and Young (1999) noted that trainees were able to utilize their group
supervision to its advantages once the group achieved more confidence and
cohesion. Emerson ( 1996) stated that creating an atmosphere of safety and trust is
needed, so that supervisees are willing to take risks in supervision to gain the
feedback necessary to learn from themselves and the other members. That
researcher believed that the supervisee's sense of professional adequacy and sense
personal adequacy is often perceived to be on the line. In both studies, the
supervisees tended to be concerned about being perceived as competent counselors.
6
Although the use of a group is widely recognized as offering more than simple
efficiency (Bernard & Goodyear, 1998~ Hayes et al., 2001 ~Holloway & Johnston,
1985; Riva & Cornish, 1995), supervisors are challenged to maximize the group's
potential. The most frequent goal for supervision is skill development and personal
growth (Hayes et al., 2001 ). There are several specific ways in which group
supervision is believed to benefit the supervisee: modeling, richer experience,
vicarious learning, and personal growth.
The group format provides supervisees with a forum to develop their
counseling skills through analysis and practice of individual skills. However, when
group supervision is used to address individual concerns one at a time, the unique
characteristics of the group setting are not utilized (Wilbur, Roberts-Wilbur, Morris,
Betz. & Hart; 1991 ). In particular, group supervision encompasses the full range of
group work, including attention to group dynamics, individual and group
development, and the general purposes of supervision (skill and case
conceptualization development, didactic training, and personal growth).
Discussion and critical analysis of specific cases is often richer and more
spontaneous when several people are participating (Westwood, 1989). Counselman
and Gumpert (1993) add that discussion of a case in a group can introduce a wider
array of perspectives and strategies than is possible in individual supervision,
because the different group supervisees typically do not see a case in the same way
and can offer a broader range of observations, thoughts, and strategies to the
supervisee presenting a case. Borders (1991) added that most supervision groups are
7
theoretically heterogeneous, and that group supervision provides practicing
counselors with opportunities to broaden their theoretical understandings. Cases
may be discussed from a multitude of theoretical perspectives, giving the supervisee
a chance to be exposed to and learn different case formulation or treatment
interventions.
Group supervisees challenge each other to consider multiple formulations and
treatment options, which promotes counselor development and cognitive complexity
(McAuliffe, 1992). Borders (1991) stated that group supervision enhances the
cognitive development of the supervisees, because they get feedback from a variety
of sources and are allowed to practice divergent thinking when the other group
members present views that are different or in conflict to their own. Group
supervision also fosters cognitive rehearsal (Hillerbrand, 1989). That is, the
supervisee continually processes, weighs, and reformulates information and
statements made by other group members.
McAuliffe (1992) stated that experienced counselors can gain help through
group supervision by having more complex and different treatment and assessment
options modeled for them from the other group members. At the same time, the
group members can challenge each other to provide more complex assessments and
treatment plans, which also promotes movement and learning. Group supervision
provides recurrent opportunities for problem solving as members test their own
perceptions and ideas against those of other group members. Not only does group
supervision provide opportunities for problem solving, but it requires active
8
participation of all members in assuming responsibility for the productivity of the
group (Hayes et al, 2001). In summary, the full exploration of alternatives by the
group is likely to lead to a more effective solution than one offered by single
members or even from individual supervision.
Kruger, Chemiss, Maher, and Leichtman (1988) suggested that both didactic
and experiential learning can occur during group supervision. For example, group
supervisors can help supervisees recognize the "here and now" (Hayes et al., 2001)
and can model how to move groups or individual clients into the here-and-now.
Group supervision also presents a unique training opportunity for supervisees to
master the individual counseling skills such as reflective listening and reflection of
feelings, as well as more advanced skills of interpretation, through the interactions
and feedback of the group.
In addition to learning these skills from their immediate experience, group
supervision also provides an ideal opportunity to learn from modeling or imitation
among peers or from the supervisor (Westwood, 1989). In this way, supervisees
may learn from the other group members or the group supervisor by watching their
interactions with each other or from the presentation of their own cases. Group
members promote each others' learning by modeling, by offering explanations of
information processing and case material, and through increasing each others'
motivation (Hillerbrand, 1992).
Modeling is more easily done in group supervision in comparison to individual
supervision because the supervisee, especially the novice supervisee, may have a
9
more difficult time understanding the expert supervisor's conceptualizations.
Hillerbrand (1989) explained that experts are not always the best models for novices
because they do not describe their cognitive processes as well, for these processes
typically have become automatic and therefore outside of their awareness.
Hypothetically, novices are more useful models for each other because novices may
communicate with each other in ways that are easier to understand, and also because
novices may be better at detecting and understanding the nonverbal cues of other
novices (Hillerbrand, 1989). Cognitive processes are made overt in language that
novices can understand and integrate. To promote modeling and skill development,
Hillerbrand (1989) suggested that supervision group leaders should encourage novice
counselors to verbalize their reasoning processes. In particular, how information was
selected, what was missing, and what decision-making rules were used. Also, when,
why, and how they used particular skills.
Group supervision also allows the other members to learn vicariously by
providing more direct exposure to other therapists and their work (Counselman &
Gumpert, 1993). The discussion often stimulates members to think about similarities
in their own cases, to formulate their own conceptualizations and strategies, and to
have a chance to provide supervision to their peers. Walter and Young (1999) found
that supervisees reported that when another student was presenting, they felt more
able to consider alternate theoretical perspectives and to "take pieces" (p. 83) from
somebody else's case and use it for their own clinical work. Supervisees were able
to listen to others' case presentations and the group discussions, and apply it to
10
themselves of their own cases. Observational, or vicarious, learning is facilitated by
supervisees' identification with the needs and goals of the other group members who
are receiving feedback or input from the group or from the supervisor (Westwood,
1989).
When supervisees are not actively discussing their own client, they can gain a
great deal from the discussion of cases. If the supervision group is functioning well,
then every member both contributes and benefits from each session.
Yet another benefit of group supervision is the learning of supervision skills
(Holmes et al, 1998). This occurs through their observations of the supervisor and
their peers in giving and receiving feedback as well as their own provision of
feedback. Borders ( 1991) also noted that "In effect, this procedure [group
supervision] provides indirect and infonnal training in supervisory skills" (p. 249).
Supervisory training for doctoral level trainees is particularly important, for many
are likely to have supervisory responsibilities in the future.
Supervisees receiving group supervision can experience personal growth.
McAuliffe ( 1992) noted that group supervision provides both challenge and support,
which he believed are fundamental conditions for development. Wilbur et al ( 1991)
found that through discussion and exploration, group members are able to enhance
and integrate their own levels of personal awareness regarding their attitudes and
beliefs. For example, when group members discuss their counter-transference issues
related to client material, it may help them to become more aware of their own
values and beliefs that are relevant to their professional work.
11
Ray and Altekruse (2000) argued that the group supervision format provides
peers and supervisors the opportunity to interact more openly and to offer support to
one another in their growth. They asserted that group supervision offers an
environment that facilitates personal growth and awareness, as well as peer support.
Westwood (1989) believes that this is because self-evaluation and self-disclosure to a
group of peers is easier than in the presence of a supervisor only. Marshall (1999)
explained that if the group manages to mirror the feelings of the supervisee, that
supervisee feels supported, validated, and freed of constraints, and becomes creative.
The group's support and understanding allows supervisees to challenge
themselves, take risks, and learn more. In their qualitative study, Christensen and
Kline (2000) found that as supervisees faced their anxiety, stayed involved with their
group supervision through interaction and participation, they became more self
aware and learned their skills more effectively. More specifically, these authors
found that as the supervisees experienced the benefits of participation, they learned
that their anxiety alerted them to possible opportunities for personal and professional
growth.
Hayes (1990) stated that group supervision allows for more accurate
perceptions of self and others through consistent feedback from others. Starling and
Baker (2000) conducted a qualitative study in which they found that the supervisees
showed an increased sense of personal identity and professional self-confidence. In
the same manner, Holmes et al. (1998) found that supervisees reported that the
support and training they received in their group supervision facilitated their comfort
12
and confidence as counselors. They believe that group supervision provided a
supportive atmosphere, and that supervisees were encouraged to test hypotheses, try
new approaches, and develop their own counselor identities. In this way, group
supervision enhances the development of both the independence and self-confidence
of the group members.
Group supervision also promotes autonomy and independence in supervisees.
Tebb et al. (1996) found that supervisees in group supervision felt that with strength
in numbers, there was an opportunity for a "rebalancing in power" (p. 43) of the
expert. That is, the group as a whole was recognized as having expertise and
knowledge, supervisees were given the opportunity to provide feedback or
interpretations, and the group members felt more comfortable questioning or
disagreeing with the supervisor. This is an important aspect of both personal and
professional growth. As supervisees continue to develop as counselors, the
atmosphere of group supervision may aid in the transition from student to colleague.
In summary, group supervision offers many benefits beyond the obvious
advantage of reducing supervisory time. There are many unique learning
opportunities to be found in group as opposed to individual supervision. However,
Tebb et al. (1996) and Bernard and Goodyear (1998) noted that whereas many
authors support the strengths of group supervision, none believe it should replace
individual supervision. Group supervision consistently has been seen as a
complement to individual supervision. Bernard and Goodyear (1998) and Hayes et
al. (2001) noted the potential disadvantages or limitations of group supervision.
13
These include time limitations, confidentiality issues, competitiveness, confonnity
pressures, unproductive feedback, and interpersonal conflict.
Studies in Group Supervision
Holloway and Johnston (1985) first reviewed the group supervision literature in
counseling training. They noted that whereas the group format is a frequent practice
in counselor training, the volume of research in group supervision is lacking.
However, these authors noted that if counselor educators continue to use group
supervision as a component in training programs, they need to make an effort to
understand its contribution and validity.
Holloway and Johnston (1985) completed a computer assisted and a manual
search of research articles in counseling psychology from 1960 to 1983 related to
group supervision. From this search, they found a total of 30 articles that addressed
group supervision specifically. Prieto (1996) continued this process and reviewed
research articles related to group supervision from 1984-1994 and found an addition
10 journal articles from this period of time. For a more detailed description of these
studies, Prieto (1996) provided a summary of those articles found in these two
reviews. As noted in the introduction, both Holloway and Johnston (1985) and
Prieto (1996) strongly believed that the lack of adequate research in the area of group
supervision was "indefensible" (Prieto, 1996, p. 305) and demanded a more
systematic examination and effort.
Prieto (1996) suggested the use of alternative methodologies to study group
supervision. He believed that isolating characteristics unique to group supervision is
14
difficult. Confounding factors such as the fact that supervisees often receive both
individual supervision and group supervision at the same time or using self-reported
perceptions of supervision processes are often difficult to relate to specific
supervisory behaviors or characteristics of group supervision. Therefore, this review
also will include findings from qualitative studies that provide useful information to
the group supervision literature.
Both Holloway and Johnston (1985) and Prieto (1996) found that early
literature in group supervision concentrated on the use of interpersonal process
groups that were prevalent during the 1960's and early 1970's. However, these
process groups were frought with ethical conflicts, for the distinction between
counseling and supervision was unclear, and also because these groups were a
mandatory component in the students' training. Prieto (1996) described this
approach as having "captive therapy ethical issues" (p. 297), but also noted that
current approaches to group supervision typically retain some emphasis on a
supportive, non-therapeutic, role for supervisors and group members. Aside from
the ethical concerns, the research failed to substantiate the efficacy of these
interpersonal process groups due to numerous methodological problems such as the
measures used, inadequate descriptions of the participants, or the methodology of the
study.
The more dominant approach in recent history, according to Holloway and
Johnston (1985), has been the case presentation approach, which is common to
individual models of supervision, which uses the supervision group to discuss trainee
15
cases and clinical issues, as well as group dynamics and didactic material.
Regarding this approach, Holloway and Johnston noted only four studies (Axelson,
1967; Blocher, 1983; Fraleigh & Buchheimer, 1969; Rioch, Coulter, & Weinberger,
1976) that provided descriptions of the case presentation format, but, again noted
that there is no substantive empirical information on this approach. This format
circumvented the ethical problems of the interpersonal process groups that
emphasized disclosure of personal material by focusing attention to the clinical work
while addressing personal awareness and growth within the professional context.
Holloway and Johnston (1985) also noted studies that emphasized
developmental approaches (Stoltenberg, 1981) to group supervision. This approach
was similar to the case presentation format in that it included components of case
conceptualization, didactic instruction, and group process. However, the content,
method of instruction, and emphasis of these components are adjusted according to
the trainees' level of professional4evelopment.
As noted previously, Holloway and Johnston (1985) had described the
literature on group supervision to be at a very "rudimentary" level of explaining and
understanding the contributions to a trainee's development. There since have been
more studies of the impact of group supervision on the supervisee's development.
Kruger et al. (1988) used a behavioral observation system and a self-designed
questionnaire to address this gap in outcome data by assessing group supervision
related to: a) supervisees' satisfaction with group supervision, b) the interpersonal
climate of group supervision, and c) supervisees' perceptions regarding their
16
knowledge and skill acquisition from their group supervision. Two studies
previously examined group supervision outcomes (Betz, 1969; McKinnon, 1969).
Kruger et al. (1988) found that supervisees, in general, were satisfied with their
group supervision meetings. Participation, total supervision time, and the
supervisees' knowledge and skill development ratings were related to higher ratings
of satisfaction in this study.
Ray and Altekruse (2000) compared the efficacy of individual supervision to
that of group supervision. In a previous study, Lanning (1971) had found no
significant differences in supervisees' perceptions of the supervisory relationship
between the two methods of supervision. Similarly, Averitt (1989) found that both
methods were equally effective in teaching empathic responding to counseling
students. In addition, Savikas, Marquart, and Supinski (1986) provided empirical
support for the effectiveness of group supervision with medical students learning
interviewing skills. Ray and Altekruse (2000) found that group supervision and
individual supervision were equally effective in increasing counselor effectiveness.
Ray and Alterkruse (2000) also found that large group size (8:1 ratio) and small
group (4:1) sizes were equally effective, and therefore concluded that there was no
empirical evidence to support any particular group size. Surprisingly, they found
that the larger group size appeared to be more effective in increasing counselor
autonomy. They believed that supervisees in larger groups may become more reliant
on themselves due to the lowered level of attention they may receive.
17
Wilbur et al. (1994) conducted a pilot study on their Structured Group
Supervision model for group supervision and found that the supervisees under this
condition reported significantly more satisfaction than those in "control" groups.
This study represented one of the first attempts by researchers to investigate the
effectiveness of a specific group supervision model. Similarly, McAuliffe (1992)
described a case presentation method of group supervision that had been developed
through its use at a college counseling center. However, no empirical support was
provided in this article.
Agnew, Vaught, Getz, and Fortune (2000) conducted a qualitative evaluation
of their peer group clinical supervision model and found that the group supervisees
observed and attributed positive counseling skill, professional gains, and personal
growth to their group supervision experience. The participants believed that the
group supervision program had increased their opportunities to learn counseling
skills and techniques, increased their sense of professionalism, and increased their
personal growth in the form of confidence, comfort with their jobs, and professional
validation. Other benefits Agnew et al. (2000) found included improved professional
relationships with their peers and the amount of support gained from their fellow
supervisees. In another qualitative analysis Marshall (1999) found five variables that
were thought to make the group function cooperatively: mutual acceptance and
support; play and creativity; tolerance and enjoyment of transference
countertransference; resolution of intragroup conflict; and gained competence.
18
Models of Group Supervision
Holloway and Johnston (1985) identified four distinct forms that supervision in
a group might take: (a) group supervision in a practicum setting of trainees who are
learning individual counseling skills, (b) group supervision of trainees learning pre
practicum interviewing skills, (c) leaderless groups in which trainees provide peer
supervision in a group format, and (d) group supervision in a practicum setting of
trainees who are learning group facilitation skills. This investigation and literature
review is primarily focused upon group supervision involving trainees receiving
supervision of their individual counseling in a group format by a clearly defined
group supervisor.
Groups are commonly classified on the basis of shared properties such as the
number of members, duration, function, membership characteristics, setting, level of
prevention, leadership styles (Hayes et al., 2001). These characteristics describe
separate, but interacting elements of the group. For example, the number of
supervisees in the group will affect the formation and development of the group
along the other dimensions. The size of the group can change the amount of
attention and feedback each member receives or affect the ability of the group to
form a sense of cohesiveness. Supervisees' experience level, or developmental level,
will affect the level and focus of supervision in that more experienced supervisees
will possess wider variability of experiences, skills, and expertise in comparison to
pre-practicum students. In addition to the properties of the group, Hayes et al.
(2001) noted that the numerous individual differences are likely to influence the
19
dynamic character of the group in supervision. Gender, racial, cultural, and sexual
orientation differences can either create complications or add to the productive
diversity of the group experience.
As previously noted, Holloway and Johnston (1985) and Prieto (1996)
observed that group supervision has historically been conceptualized and
implemented from two different approaches: the interpersonal process approach and
the case presentation approach. In addition, the developmental approach also has
been described in the literature, as have the interpersonal process approaches that
focus on increasing supervisee self-awareness or emotional growth. This latter
model closely resembled group counseling in many ways, and therefore was fraught
with ethical and professional issues (Holloway & Johnston, 1985). Holloway and
Johnston (1985) describe the case presentation approach as being similar to
individual models of supervision in that supervisees presented clinical material, but
that at the same time attention was paid to group process and personal awareness and
growth within a professional context. The developmental approaches to group
supervision consider the supervisees' level of development as counselors. These
models are similar to the case presentation format in that they include components of
didactic instruction, group process, and case conceptualization, but also include
general strategies for trainees at different levels of professional development.
Both Holloway and Johnston (1985) and Prieto (1996) criticized the group
supervision literature for the lack of research and models. Preito (1996) stated,
"Unfortunately, there are neither adequate models nor convincing empirical studies"
20
(p. 338). A search of the literature revealed a handful of models for group
supervision that will be briefly described and summarized in this section. However,
several of these models are described theoretically or are in use in different training
facilities, with the findings reported anecdotally or through smaller pilot studies.
In a model called the "Case Presentation Approach" McAuliffe (1992)
described an approach to group supervision that has been employed in a college
counseling setting for several years. This model involves case presentations in
which the supervisee utilizes the familiar SOAP format (subjective, objective,
assessment, and plan) and the group members provide support and feedback in these
various domains. The group supervisor plays the role of process facilitator and
expert.
Wilbur et al. (1991) presented what they called the "Structured Group
Supervision" (SGS) model. Their model was structured to encourage the active
involvement and participation of all group members, rather than having the
supervisees take turns receiving supervision from the supervisor. Their model
incorporated five phases: the request-for-assistance statement, the question period
and identification of focus, the feedback statements, the supervisee's response to
feedback, and then a discussion period. These phases are described in more detail in
Wilbur et al. 's (1991) article, but essentially provides a format and structure for
group supervisors to help ensure interaction and participation from the group
members.
21
A "Systematic Approach to Peer Group Supervision" (Borders, 1991)
encouraged skill development, conceptual growth, participation, instructive
feedback, and self-monitoring. The supervisor functioned as a moderator and
process observer. A unique aspect of this model is that the other group supervisees
periodically were asked to assume certain "roles" such as client, counselor, outsider,
etc. to help the supervisee broaden their perspective of their clients. In addition,
Borders (1991) provided guidelines for focused observations, increasing theoretical
perspectives, and using metaphors.
Based on her clinical experience as a supervisor, Emerson (1996) presented a
model for group supervision that was found to increase supervisee self-esteem,
willingness to take risks, genuineness, and better clinical work. She stressed group
processes focusing on cooperation, responsibility, relatedness, and respect. Her
model included a set of norms and rules for attentive listening, constructive
feedback, voluntary participation and disclosure, and confidentiality.
Integrating several models of individual and group supervision, Agnew et al.
(2000) presented a group supervision model that borrowed elements from Bernard's
Discrimination Model (1979), Systematic Peer Group Supervision (Borders, 1991),
and Wilbur et al. 's Structured Group Supervision ( 1994) approach. In this model,
when presenting to the group, the supervisees briefly describe the client, stating the
counseling goals and desired outcomes, the kind of help wanted, and the desired
supervision method. The group feedback included strengths of the supervisee
22
comments about the case, and suggestions for improvement, for strategies, and for
direction.
Using concepts borrowed from object relations theory and group dynamics,
Altfeld (1999) presented an experiential group supervision model. The central
component in this model is that material presented in group supervision stimulates
conscious and unconscious parallel processes in the group members. These
countertransference issues that may have eluded the presenter can become evident
through the reactions of the other group members, and then can be identified and
worked through on an emotional and cognitive level. Moss (1995) outlined a similar
group supervision model in which the group may be used to reflect, amplify, and
process the presenting supervisee's countertransference experience. However, Moss
(1995) also warns that a focus on countertransference issues may blur the boundaries
between supervision and therapy, and that these sorts of exercises should always be
intended for the benefit of the client rather than the personal needs of the supervisee.
The Role of the Group Supervisor
Individual and group supervision involves teaching specific clinical skills, the
feedback of observations, and the personal development of the trainee. Group
supervision represents a level of supervision that presumes prior experience and
demonstrated mastery in individual supervision. Kruger et al. (1988) argued that
group supervision may be a more complex task than individual supervision, and that
the novice supervisors may be at a greater disadvantage in this context.
23
In this section research from both the group supervision and group therapy
literature have been combined to attempt to further describe different aspects of the
group supervisor. Counselman and Gumpert (1993) stated, "we take the position
that group supervision is not individual supervision in the presence of a few other
people" (p. 30 ). Similarly, according to Hayes et al. (200 I), being the supervisor of
counselors in a group is not the same as supervising them individually. The presence
of a supervisor in a group can create a pull toward a model in which the leader does
individual supervision in front of the other members. According to Counselman and
Gumpert (1993), group supervision has "an important synergistic effect, but only if
the leader avoids dyadic supervision" (p. 30). They believe that more and possibly
superior creative insights into a therapeutic case can be generated by the supervisor
plus the group, than by the supervisor alone.
The effective group supervisor must, first and foremost, be an effective group
leader. Hayes et al. (200 1) stated that "the essential task of the group supervisor is to
facilitate the development of a productive work group before effective supervision
can begin" (p186). McAuliffe (1992) noted thatthe role of group supervisor is a
combination of process facilitator and expert. The supervisor provides modeling,
diagnostic expertise, and knowledge of group dynamics. Hayes et al. (200 1) asserted
that the supervisor should not act so much to direct the group as to create a climate in
which the group finds its own direction. At times, it may be more important that the
supervisor give the group the opportunity to work with a case or issue under
24
consideration, than it is for the supervisor to dispense his or her own knowledge or
experience first, which could hinder the group interaction.
Whereas McAuliffe (1992) also noted that the presence of a supervisor with
more extensive experience prevents the danger of "the blind leading the blind" which
may occur in group situations in which the only members are peers, Counselman and
Gumpert (1993) suggested that the teaching function should be secondary to the
group supervisor's function as the group facilitator. The group supervisor's role
consists of a "balance between non-directively promoting counselor problem solving
and more directively offering insight and even suggestions" (McAuliffe, 1992, p.
165). That is, promoting the supervisee's attempts to problem solve and
conceptualize a case, while at the same time offering more concrete or didactic
information when needed.
Borders ( 1991) stated that the two critical roles of the group supervisor are that
of moderator and process observer. She stated that the supervisor has the
responsibility to help the group stay on-task; sets up practice exercises, directs skill
practice, and summarizes feedback and identifies themes for groups. In addition to
this, the supervisor serves as a process observer of group dynamics: describing
patterns of interaction and encouraging discussion of behaviors, feelings, and
relationships. In this way, the group supervisor must be sensitive to group dynamics
that occur.
In addition, Borders (1991) cautioned that group supervision sessions may not
necessarily be helpful or productive. For example, peers may be overly supportive
25
and prone to giving advice, or the group may experience difficulty staying focused or
on task. To be effective, Borders (1991) stated that group supervision needs an
organizational structure and group members need training in supervisory skills, and
the group supervisor must maintain responsibility for fostering constructive, helpful
feedback and keeping the group focused and on task.
Counselman and Gumpert (1993) believe that the role of the group supervisor
is to create and maintain a safe, growth-enhancing group environment through
careful attendance to group dynamic issues that may interfere with the group's
ability to focus on the task of supervision. They suggested that the group
supervisor's skills should be based on well-established principles of group
facilitation from the fields of group therapy and group development that can be
applied to group supervision.
The style that a group leader displays has a direct effect on the behavior of the
group members. Hansen, Warner, and Smith (1980) stated that most models of group
leadership have been variations of Lewin's (1944) concepts. Lewin identified three
basic styles of group leadership: authoritarian; democratic; and, laissez-faire.
Authoritarian group leaders envision themselves as experts, interpret, give advice,
and generally direct the movement of the group. These leaders often equate group
leadership with instructing. Democratic group leaders are more group centered or
nondirective. Similar to Rogerian theorists, these leaders trust the group members to
develop their own potential and that of the other group members. Democratic
leaders serve as facilitators of the group process and not as directors of it (Gladding,
26
1991). Finally, laissez-fair leaders are those who fail to provide any structure or
direction for their groups. The members are left with the responsibility of leading
and directing. This type of leader is often attempting to be nondirective, but fails to
facilitate the group in any way. Gladding (1991) notes that these types of groups
often are unproductive, or in the worse case, the group may become out of hand and
do damage to its members.
Regarding group leadership in therapy groups, Wheelan (1997) asserted, "the
hunt for the characteristics of effective leaders, however, has been long and generally
unproductive" (p. 290) and concluded that different leadership styles are needed at
different stages of group development. Gladding (1991) notes that most effective
leaders show versatility and modify their leadership pattern to coincide with the
purpose of the group and its group members. Wheelan (1997) added that leaders
alone cannot facilitate group movement or ensure group or individual goal
achievement without the active participation of the group members in the process as
well.
Strategies for Group Supervisors
McAuliffe (1992) believed that one of the group supervisor's major tasks is to
promote divergent thinking and problem solving by encouraging active learning,
interaction among group members, and reasoning. The supervisor also focuses on
affective development, specifically promoting the supervisees' increased awareness
of their attitudes and feelings towards their clients, themselves, and each other.
27
Supervisors need group leadership skills to help build a supportive envirorurient
of appropriate relationships and communication between members (Blocher, 1983).
For this reason, the supervisor must be especially alert to group dynamics, which can
help move the group toward greater exploration, openness, and responsibility. Hayes
et al. (2001) noted that creating an effective group experience depends on the
supervisor's ability to provide both high levels of positive interdependence for
achieving a common goal as well as a group identity that unites the diverse members
based on several different sets of values.
In a study of medical students receiving group supervision, Savikas et al.
(1986) found that students judged the following role requirements to be most
important for group supervisors: 1) modeling target behaviors; teaching skills,
techniques and strategies; 2) facilitating exploration, critical thought, and
experimentation; and 3) evaluating performance. Although Savikas et al.'s (1986)
study was conducted on medical students, those authors believed that counseling
students behave in a similar way in practicums that involve generic skill training.
Regarding the performance of the group, Wilbur et al. (1991) stated that
increased structure may improve group effectiveness. They noted that a high level of
leader-imposed structure tends to enhance interpersonal behavior during early stages
of group development. Wheelan (1997) suggested that the initial sessions be devoted
to allowing the group members to discuss their goals and expectations, make explicit
the group guidelines and rules, and also educate the members about group process
and group dynamics so that they can help facilitate the group's development.
28
Y alom (1985) had stated that group cohesion is an important determinant for
therapeutic change in groups, leading Caple and Cox (1989) to explore the
relationship between group structure and the development of group cohesion. They
theorized that too much ambiguity in the early stages of the group resulting from
lack of clarity about the groups' goals, processes, and expectations may create
anxiety and frustration among the group members, which would result in less
cohesion and less therapeutic productivity. Similarly, Walter and Young (1999), in a
qualitative study of group supervision with social work students, found that the
supervisees felt that the "public exposure of the group supervision experience
initially amplified their anxiety" (p. 87). In another qualitative study, Christensen
and Kline (2000) found that supervisees were anxious about how they would be
perceived by other group members and by the group supervisor, and that this in tum
affected the way they participated in the supervision group. Similarly, Emerson
(1996) noted that novice supervisees are often initially fearful of their evaluation by
the supervisor and the other group members.
Caple and Cox (1989) found that providing initial structure in the form of
simple exercises resulted in the development of group cohesion among the
participants. They believed that the initial structure may have improved
communication and reduced more quickly any ambiguity and anxiety felt by the
group members. However, as the group becomes more cohesive and trusting, the
positive effects of structure decreases and that the continued use of high structure
may impede the group process (Wheelan, 1997; Wilbur et al., 1991). Therefore, as
29
the level of skill and personal development of the supervisees increase and they
become more trusting and cohesive as a group, less leader intervention and group
structure is necessary.
As supervisees become more competent and the group more cohesive, the
supervisor should shift the responsibility for group structure to the presenting
supervisee and the group members (Ellis & Douce, 1994). In a qualitative analysis
of group supervision, McAuliffe (1992) also found that the supervisors took a more
active, didactic role at first. But that later on, the supervisor became less active,
using probe questions, making process observations, and only contributing
occasional diagnostic impressions.
Holloway and Johnston (1985) found that most supervision groups involved
three domains of focus: didactic; case conceptualization; and, interpersonal process.
Wilbur et al. (1994) identified three categories of supervision groups: the task
process group modality; the psycho-process modality; and, the socio-proces
modality. The task process group is a combination of didactic and case
conceptualization material. The psycho-process modality parallels the intrapsychic
growth expected in the interpersonal process group. Finally, the socio-process
modality parallels the interpersonal relationship growth expected in the interpersonal
process group.
Wilbur et al. (1991) noted that whereas group supervision may focus on case
presentations, didactic material, or interpersonal processes, it is appropriate for the
supervisor to override the supervisee or the group's focus when it appears that the
30
group is engaged in resistance. For example, a group may be presented with some
difficult counter-transference or interpersonal process issues, and then choose to
focus on clinical material to actively avoid discussion of the more difficult subject
matter. In cases such as this, Wilbur et al. (1991) suggest that the supervisor model
appropriate counseling skills by refocusing the group.
In summary, Westwood (1989) provided the following suggestions for the
group supervisor: (a) establish group norms regarding confidentiality and
communication, (b) establish group cohesion in the initial stages, (c) set individual
goals within the group, (d) model all the skills that the supervisees are required to
use, and (d) apply the essential group leadership skills to effectively manage the
group processes.
Strategies Related to the Developmental Level of the Supervisee
Another question that arises when considering the use of group supervision is
the extent to which the supervisee training level relates to the process of group
supervision and their needs and expectations of the group supervisor. Since
Holloway and Johnston's ( 1984) first review of the group supervision literature,
there has been a growing interest in developmental models in supervision that
explore how supervisors and supervisees change as they gain experience. Although
many authors have written about the developmental level of the supervisee as it
relates to individual supervision, there is not as much literature that discusses the use
of group supervision in relation to the training level of the supervisee.
31
Many theories of counselor development assume that counselors at different
levels of experience have different supervisory needs (Swanson & O'Saben, 1993).
For supervision to be effective, satisfying, or growth promoting, supervisors must be
able to meet the needs and expectations of their supervisees at all experience levels.
Savickas et al. (1986) suggested that the level of training may be a more important
variable in research on effective group supervision than is type of student.
Hayes (1990) encouraged group supervisors to consider the process of group
supervision as occurring in three different but interacting domains, each with
different levels of development. The group supervisor must consider the level of
cognitive complexity of the supervisee, the developmental level of the group, the
level of training of the group members, and the interactive effects of these variables
with one another. That is, different levels of individual development interacting
across different stages of group development require different leadership responses.
For example, novice counselors in the working stage of a group would need a
different amount of input, feedback, or guidance from the group supervisor than
experienced supervisees. Stoltenberg's (1981) suggested that supervisees at this
level lack experience, confidence, and awareness of their impact on clients. They
tend to want advice and directions from their supervisors and to seek the correct
ways to behave. Swanson and O'Saben (1993) found that supervisees with less
practicum experience expressed greater need for supervision that focus on concrete
skills and facts and with supervisors who take a more active and directive role in the
supervision process.
32
Similarly, Borders (1991) noted that the group supervisor must be cognizant of
the developmental level of the group members and that for novice supervisees, the
group supervisor may be more active and directive by taking more responsibility for
assigning tasks for members, eliciting feedback, and summarizing discussions.
While for more advanced counselors, Borders (1991) encouraged the group
supervisor to allow the supervisees to assume more responsibility for the conduct of
the sessions, to be more verbal about requesting feedback and opinions, and even to
address group process concerns.
Likewise, Heppner and Roehlke ( 1984) found that novice counselors rated
supervision as better when their supervisors provided support as well as skill
acquisition and training, whereas more advanced doctoral interns favored
supervision that examined personal issues affecting therapy. Heppner and Roehlke
(1984) concluded that supervision needs appeared to follow a progression from
support/awareness/enhancement issues to more self-disclosing, personally
threatening types of issues as the supervisees progressed through their own
development. That is, novice counselors want techniques and directions on how to
do counseling, whereas more experienced supervisees become interested in
examining personal issues that relate to their work.
In a review of the literature related to counselor development Stoltenberg,
McNeil, and Crethar (1994) concluded that beginners valued support and technical
direction, while advanced trainees found personal issues and relationships affecting
their professional work more important. Heppner and Roehlke ( 1984) found that
33
across supervisee developmental levels there were no significant differences for
either general or specific expectations about the supervisor or of themselves as
supervisees. However, they believed that supervisee perceptions of supervisor
expertness, attractiveness, and trustworthiness change as the supervisee becomes
more sophisticated. For example, doctoral interns may attend more closely to what
is said, rather than who is saying it.
Specific Skills of the Group Supervisor
Harville, Mason, and Jacobs (1983) wrote 20 years ago that there had been
very little research on group leader training or even on the specific skills necessary to
lead groups. They believed that the field of individual counseling is further ahead in
terms of training approaches and a number of counseling skills that are
therapeutically positive. Nearly a decade later, Gladding (1991) also stated, "group
research is in its infancy when compared to research in the other social sciences" (p.
335) and he noted that one of the primary reasons is the complexity of studying
groups. Studying dyadic relationships in individual counseling is more simple in
comparison to the interactions and impact that group members have on each other.
A second reason that research is difficult is the lack of process or outcome
instruments related specifically to groups. Gladding (1991) noted that exceptions
were the Group Attitude Scale (Evans & Jarvis, 1986), the Group Climate
Questionnaire (MacKenzie, 1983), the FIRO-B (Schutz, 1958), and the Group
Satisfaction Scale (Caple & Cox, 1989) that focus on group dynamics and outcome.
34
Gladding (1991) summarized many of the counseling skills necessary to
facilitate a group. These included skills that have been borrowed from individual
counseling and are not unique to the group setting. Skills such as active listening,
clarifying, summarizing, questioning, interpreting, confronting, reflecting feelings,
etc. are used constantly when facilitating a group. However, Harville et al. (1983)
note that there are more specific skills needed for leading a group, whether these are
therapy groups, supervision groups, discussion groups, or staff conferences.
Gladding (1991) and Harville et al. (1983) identified eight basic skills needed to
facilitate a group. These include redirecting the conversation, eliciting the
participation of members, focusing on certain issues or topics, paying attention to
nonverbal cues, connecting members experiences, protecting and supporting
individuals, and reality testing.
Group development over time
Yalom (1995) suggested that, although each group is unique, effective small
groups follow a generalizable pattern from initiation to termination. Wheelan (1997)
summarized that in all group development models, groups are seen as moving
through a phase focused on creating a sense of safety and membership, followed by a
period with the emergence of conflict among members and between members and
the leader. Next, the group concentrates on deepening interpersonal relationships
and organizing more effectively to accomplish tasks. The following stage is a period
of intense focus on work and goal accomplishment, which is then followed by a
phase of termination near the group's end. Although the particular names and
35
boundaries of these stages vary from description to description, groups seem to
follow a similar general sequence (Tuckman & Jensen, 1971 ).
In two qualitative studies, Walter and Young (1999) and Christensen and Kline
(2000) found that in the early stages of supervision groups, the supervisees tended to
feel anxiety related to their concerns of how they would be perceived by the other
group members and the supervisor. The anxiety restrained their involvement and
blocked growth. Holmes et al. ( 1998) found that in the early stages of group
development, it was important for supervisors to demonstrate genuine interest and a
nonjudgmental or problem-solving stance to aid in building a foundation for work in
the middle and final stages. They believed that the supervisory alliance was more
important to establish in these early stages.
Christensen and Kline (2000) noted that in their study as the group supervisor
continued to encourage interaction and as the feedback exchanges increased, the
supervisees began to realize that the others were not evaluating them, but were
offering helpful observations and constructive feedback. As the interactions
continued, trust and cohesion developed, and supervisees became more convinced
that their peers and supervisor were invested in providing feedback that could
facilitate learning and growth. Holmes et al. (1998) found that during the middle
stages of group supervision, the supervisees felt increasingly more comfortable in
disclosing personal issues related to their training and clinical experiences,
discussing group dynamics, countertransference, and ethical and professional
dilemmas.
36
In their study of supervision groups, Kruger et al. (1988) found that the
problem-solving behavior (performing) of the groups occurred in an "oscillating
pattern of increases and decreases over time" (p. 615). They reasoned that the
groups' focus on task-related issues builds up a reservoir of group tension that must
be released through involvement in non-task-related matters, and that during the
performing stage one might expect an oscillating equilibrium such as this.
Wheelan ( 1997) believed that the first stage of group development typically
appears to last at least two months, with many groups requiring more time to move
through this stage. The conflict stage has never been identified in a work context
earlier than two months into the life of the group, and the work phase occurs
approximately six month's from a group's beginning (Wheelan, 1997).
Purpose of this Study
The purpose of this study was to develop and provide preliminary validation
evidence for the Group Supervision Scale (GSS}, a measure to assess a supervisee's
perception of the group supervisor. Items were developed using focus groups
consisting of graduate student supervisees at the predoctoral intern level and expert
supervisors. After the scale was developed, the investigator engaged in a multi-step
process to validate the scale. The first step used an exploratory factor analysis to
determine the number of dimensions or factors involved in the evaluation of a group
supervisor. The second step focused upon measuring how the performance of the
group supervisor related to other constructs associated with group supervision. In
particular, to provide evidence for convergent validity the constructs of supervisee's
37
perception of group climate, supervisee satisfaction with supervision, and the
emphasis the supervisors placed on promoting the trainee's understanding of the
client and development of counseling skills were analyzed with regard to their
relationship to the GSS. In addition, the GSS was compared to the supervisor's age,
level of physical attractiveness, and appearance professionally to determine the
discriminant validity of the instrument.
Group Climate
In a review of the group supervision literature Prieto (1996) noted that the
interactions among trainees seem to be a primary and highly influential focus of
group supervision. Despite their characterization by such static dimensions as size,
membership, duration, purpose, and so forth, groups have a dynamic quality to them.
Hayes et al. (2001) noted that group is always moving, doing something, changing,
becoming, interacting, and reacting. Group dynamics are created by the struggle by
group members to balance the forces associated with accomplishing goal-related
tasks and building a shared community.
Two possible interpersonal processes that occur during group supervision were
noted by Bernard and Goodyear (1998): between-supervisee competition and
between-supervisee support. Most supervisees find group supervision to offer a
greater amount of support than individual supervision. The process of group
cohesiveness occurs as the individual members of the group begin to invest
emotionally in the group and with each other. Initial cohesion often becomes the
basis for group support and the development of trust. Greater trust encourages
38
cooperation which has been shown to increase productivity and achievement in
groups with set goals (Westwood, 1989). This leads to an atmosphere of support and
a higher level of trust, which allows the supervisees to become more willing to
openly discuss their clinical work.
Kruger et al. (1988) found that supervision groups with the highest levels of
cohesion were groups that were observed to have the most frequent "discussions of
counselor problems" (p. 615). They, however, stated clearly that it was impossible
to determine if the discussion of counselor problems might improve the cohesion of
the group, or if the cohesion facilitated the discussion of counselor problems.
However, their findings also support those stated above, that group members are
more willing to participate and discuss their clinical work, countertransference, or
other relevant issues when there is more cohesion in the group.
Group cohesion appears to contribute to a number of important variables.
Evans and Jarvis (1986) stated that attraction to group has been equated with group
cohesiveness in that members who find their group attractive are more likely to
remain members of the group, to attend regularly, and are more willing to contribute
to group discussion and self-exploration. That is, group cohesiveness is believed to
increase participation and engagement of the group members.
In the same way, Wheeler and Kivlighan (1995) believed that lack of verbal
participation was related to dissatisfaction in the group. A number of other studies
suggest that the amount of group members' verbal participation is related to
enhanced group cohesion (Caple & Cox, 1989; Emerson, 1996; Yalom, 1985).
39
Increased verbal participation is thought to enhance cohesion by increased
interaction and self-disclosure.
The group supervisor should attempt to establish and maintain group cohesion,
especially early in the group history. Supervisors may facilitate group cohesion
through structured group activities, connecting members, or pointing out similarities
among members (Hayes et al., 2001). Supervisors can inhibit the development of
group cohesion by remaining in the role of expert in the group and treating the group
supervision as a session in which each member gets individual guidance, rather than
facilitating group participation.
Counselman and Gumpert (1993) warn that personality conflicts inevitably
occur, and that it is important that the group attempt to resolve these conflicts.
Competition often develops among the members in a supervision group (Ellis &
Douce, 1994) for a variety of reasons. In group supervision, the supervisees' needs
to feel and to be perceived as competent may fuel between-supervisee competition
(Goodyear & Bernard, 1998). Supervisees with more experience may become
condescending or may tout their respective skills. AUhough this competitive
atmosphere may serve as useful in terms of stimulating a group, it also may affect
the group in a negative way by creating conflict, envy, or withdrawal.
There are other sources of conflict as well. Individual differences in race,
culture, gender, or sexual orientation may force individuals to try to isolate
themselves from the group or group process, or become ostracized for being different
(Ellis & Douce, 1994). As noted above, Counselman and Gumpert (1993) stated the
40
importance of the group to attempt to resolve these types of conflict. Discussions are
important for the development of the group as a whole, and the development of the
individual supervisees. In addition to personal growth, addressing these conflicts
provide a model of here-and-now work that is valuable as a counseling skill.
The first hypothesis in this study was that a high rating on the Group
Supervision Scale would be significantly related to high levels of group cohesion.
Although a group's cohesiveness is not solely dependent on the group supervisor,
this author hypothesized that group supervisors with higher ratings will be able to
foster higher group cohesion through their facilitation skiils. Group cohesion was
measured by the Group Climate Measure (MacKenzie, 1981 ).
Satisfaction with Supervision
Supervisee satisfaction with supervision may be defined as the supervisee's
perception of the overall quality of supervision, and the extent to which supervision
met the needs and facilitated the growth of the supervisee (Ladany et al. 1996).
Heppner and Handley ( 1981) the importance of supervisee satisfaction is based on
the belief that satisfaction is necessary for the supervisee to be motivated to work to
achieve his or her goals in supervision.
Kruger et al ( 1988) found a positive correlation exists between the amount of
participation during group supervision sessions and the degree of satisfaction. They
found that supervisees who participated more during group supervision tended to
report more satisfaction, and also that non-participation was negatively correlated to
satisfaction. The participation of the group members may be dependent on several
41
factors such as the group environment (e.g. cohesion, trust, safety, etc.) and also the
group supervisor's ability to facilitate participation. Harville et al. 's ( 1983) list of
basic skills illustrates the techniques and strategies the group supervisor may use to
facilitate participation and discussion by all the group members.
Therefore, the second hypothesis in this investigation was that there would be
a significant positive relationship with the supervisees' perception the group
supervisor and their satisfaction with supervision. The Supervision Satisfaction
Questionnaire (Ladany et al., 1996) was used to assess this variable.
Skill and Conceptualization Development
Promoting the trainee • s understanding of the client and development of
counseling skills is arguably the most important aspect of supervision, whether it is
in the individual or group format. Watkins (1997) stated that clinical supervision is
important because it provides supervisees with feedback about their performance,
and gives them alternate views and perspectives about patient dynamics,
interventions, and courses of treatment.
Individual and group supervision enhances the professional functioning of the
counselor. Supervision helps the therapist become better and more effective with
regard to conceptual ability, intervention, assessment, and implementation (Watkins,
1997). Bernard's ( 1979) Discrimination Model of supervision proposed that
supervision consists of three functions: to develop the counselor trainee's process,
conceptualization, and personalization skills. According to Bernard ( 1997),
"process," or intervention, skills include all of those behaviors that distinguish
42
counseling as a purposeful therapeutic interpersonal activity such as empathizing,
confrontation, and interpretation, among others. A supervisee's conceptualization
skills includes the ability to make some sense of the information that the client is
presenting, to identifY themes, and to discriminate important information from non
important. The third function from Bernard's model is analogous to the term
''personal growth" and refers to the contribution of the supervisee as an individual.
According to Wampold and Halloway (1997), the goal of supervision is to
change some characteristic of the therapist, such as their skill level, which will in
turn result in more competent delivery of counseling. These authors believe that one
ofthe main goals of supervision is to increase the skill level ofthe supervisee. By
working with the supervisee on specific clients, not only does supervision affect the
client directly, but these skills and knowledge help the supervisee recognize similar
therapeutic situations in the future.
In summary, group supervision should promote the development of
counseling skills and enhance the supervisee's ability to understand a client.
Therefore, the third hypothesis of this study was that a high rating on the Group
Supervision Scale would be significantly related to high scores on the "client focus"
subscale of the Supervisory Working Alliance Inventory (Efstation, Patton, &
Kardash, 1990). This subscale reflects the level of emphasis that supervisors place
on promoting the trainees' understanding of the client and the development of
counseling skills.
43
Age. Physical Attractiveness. and Professional Appearance
In order to establish the discriminant validity of the GSS. the scores on the GSS
was compared to the age of the supervisor, and ratings of the supervisor's level of
physical attractiveness and professional appearance. There is no conceptual or
empirical reason to believe that these variables would be related to perceptions or
ratings of the group supervision. Therefore, the fourth hypothesis of this study was
that scores on the Group Supervision Scale would not be related to the supervisor's
age, level of physical attraction, or appearance professionally.
SwnmmyandHypotheses
Validity Hypothesis I: A high rating on the Group Supervision Scale would be
significantly related to high levels of group cohesion as reflected in the Group
Climate Measure (MacKenzie, 1981).
Validity Hypothesis 2: Higher ratings of the GSS would be significantly related
to satisfaction in supervision as measure by the Supervision Satisfaction
Questionnaire (Ladany et al., 1994).
Validity Hypothesis 3: Scores on the GSS reflecting skill development and case
conceptualization would be significantly related to the "client focus" scale of the
Supervision Working Alliance Inventory (Efstation et al.. 1990).
Validity Hypothesis 4: Scores on the GSS would be independent of the
supervisor's age, level of physical attraction. and appearance professionally.
44
Chapter II
Method
The development and validation ofthe Group Supervision Scale
encompassed two studies, one building on the other. This chapter will describe the
participants and procedures followed to complete both of these studies.
Study I entailed two phases, and focused on item development to accurately
and fully represent the different aspects of group supervision. The first phase
involved two different focus groups consisting or predoctorallevel interns from two
university counseling centers to create items for the instrument. The second phase
involved having expert raters analyze the proposed items. Study 2 entailed
administering the scale to a national sample to develop norms and collect evidence of
validity.
Study I
Phase 1: Item Development
The purpose of this study was to develop the items of the Group Supervision
Scale.
Participants. The first focus groups of 4 predoctoral intern level counseling
psychology students were used to develop a pool of items. Three of these graduate
students were predoctoral interns at a university counseling center, and one had
already completed her internship and was in the process of completing her
dissertation. This focus group consisted of I white male. I white female, I Asian
American female. and I Mexican-American female. The interns ranged in age from
45
30 to 35 years old (M=33.50, SD=2.38). Each intern had an average of 6. 75
semesters (SD = 1.89) of experience receiving group supervision. and 8.25 semesters
of practicum experience (SU = 1.26).
A second focus group consisting of 4 predoctoral intern level graduate students
(different from the first focus group) in counseling and clinical psychology were
asked to review and edit the existing items. All 4 interns were predoctoral interns at
a university counseling center at that time. This focus group consisted of 1 White
male, 1 Black female. 1 White female, and 1 Asian female. The group ranged in age
from 28 to 33 years old (M = 30.50, SD = 2.08). Each intern from this second group
had an average of9.75 semesters (SD =.50) of experience receiving group
supervision. and 11.50 semesters (SD = 1.00) ofpracticum experience.
Procedure. The first group met one time for two hours at their workplace. The
author asked the focus group to create items "to assess the performance of a group
supervisor." As an example, the author showed the group members the Supervisory
Working Alliance Inventory (Efstation et al., 1990). The author instructed the focus
group to concentrate on domains that are more specific to group supervision in
comparison to individual supervision. The focus group members began by
discussing what they believed were the most important aspects of group supervision.
At times. the investigator prompted the group members by asking them "what would
the supervisor do to foster [these aspects]?" and "How could this be reflected as
items for an evaluation instrument?" One of the group members listed down the
items created by the group during their discussion. At the end of the focus group.
46
items that appeared to be redundant were eliminated by the group. From this group,
31 items were compiled.
The author met with a second group for two hours at their workplace. These
group members were asked to review these 31 items, and if they thought any
important aspect of evaluating a group supervisor was missing from the scale or if
any items should be deleted. Based on their input, items that were thought to be
unnecessary or redundant were eliminated and several additional items were added.
After this second focus group, a pool of 24 items were kept.
Phase 2: Expert Raters
This second phase of Study 1 was to evaluate content representativeness and
relevance of the items to the area of group supervision. The purpose of this phase
was to develop content validity evidence.
Participants: Four expert raters were then recruited to review the items. These
raters are considered to be "expert" based on Jennings and Skovholdt's (1999)
criteria for expert counselors. These raters were considered to be experts on the
basis of their years oflicensure (M = 15.25, SD = 5. 74), years of experience
conducting group supervision (M= 10.00, SD = 5.01), and years providing
individual supervision (M = 14.25, SD = 7.37), and by peer selection. The four
experts consisted of 2 White males, I White female, and I Asian-American female.
These experts were recommended by professors and supervisors of this author, based
on their amount of experience in providing individual and group supervision and
their work as a licensed clinician. The investigator contacted 7 potential experts via
47
electronic mail, and sent them cover letters explaining the purpose of the study and
the nature of their desired participation. Two of the individuals did not meet the
minimum criteria of 10 years of licensure and 5 years of experience providing group
supervision.
Procedure: The raters were provided with a brief review detailing
characteristics of effective group supervision and a form that contained all the
preliminary items of the GSS. For each GSS items, the raters were asked "to what
extent does this item reflect a component of group supervision." Each item was
rated on a 5-point scale ranging from strongly disagree (1) to strongly agree (5).
Results indicate that the expert raters generally were in agreement about which
items represented important aspects of group supervision. Means for the 24 items
were as follows: 4.50, 3.50, 5.00, 4.25, 3.50, 4.00, 4.00, 4.25, 4.25, 3.50, 4.25, 2.75,
3.50, 4.75, 5.00, 3.75, 4.00, 4.25, 4.50, 4.25, 4.50, 4.00, 4.25, and 4.00. All of these
means were above the midpoint of the 5-point scale. However, this author chose to
delete those 8 items that did not receive an average rating of 4.00 or higher. The
expert raters were also asked if they thought any important aspect of group
supervision was missing from the scale or whether or not any item should be deleted.
Changes in the wording of some items were made as a result of suggestions by the
raters. It was hoped that using the expert raters in this way would help improve the
content and ensure relevance of the items. From their feedback and the results of the
item ratings, a final pool of 16 items were kept that comprise the Group Supervision
Scale.
48
The goal of the third phase of the study was to develop norms and collect
evidence of construct validity for the GSS through several evidential procedures. It
was estimated that a sample size of 160 would be needed to validate the scale.
Because the GSS consisted of 16 items, a sample size of 160 would provide an
adequate number of participants to variable ratio (10:1) based on Grimm and
Yamold's (1998) recommendations.
Study 2
The goal of Study 2 was to extend Study 1 by developing norms and
validation evidence of the GSS through a national sample of supervisees.
Participants. It was estimated that a sample size of 160 would be needed to
validate the scale. This sample size was arrived at using the participant-to-item
ration recommended by Bryant and Y arnold (1995) to adequately complete a factor
analysis.
Participants were 227 trainees (180 women and 47 men; 180 White, 5
African-American, 15 Asian-American, 13 Hispanic, 3 Native American, and 11
biracial/multiracial) averaging 30.39 years of age (SD = 6.76). Participants were
largely enrolled in counseling psychology (72%) or clinical psychology (28%)
programs and were pursuing master's (14%), Ph.D.'s (69%), or Psy.D. (17%)
degrees. The participants had received group supervision in university based
practicum courses (41 %), university counseling centers (40%), community agencies
(16%), and hospital/inpatient settings (3%). Participants averaged 5 semesters of
supervised practica (SD = 3.02). Twenty three were 1st year students, 63 were 2nd
49
year students, 48 were 3rd year students, 30 were 4th year students, 37 were 5th year
students, and 23 were 6th or more year students. Participants had been involved with
supervision groups that had been meeting for an average of 15 weeks (SD = 10.44)
with an average of 5 supervisees in each group (SD = 2.17).
Their supervisors: Forty nine percent of the trainees' supervisors were
women. fu terms of ethnicity, 82% were White, 8% Asian-American, 7% Hispanic,
2% African-American, and 1% biracial/multiracial. The average estimated age of
the supervisors was 43.61 (SD = 8.48). On a 6-point Likert-type scale, the trainees
rated their supervisors physical attractiveness (M = 3.82, SD = .97) and professional
appearance (M = 4.62, SD = 1.12) for purposes of discriminant validity.
Measures
The purpose of the study was to develop and validate the Group Supervision
Scale. Its psychometric properties and indicator of validity are described in the
Results chapter. Four measures were used for validations purposes and to describe
the sample. These measures will be described in this section.
Demographic questionnaire. The demographic questionnaire was designed to
gather information on participants' age, gender, ethnicity, field, degree program,
number of semesters of completed practicum, year in graduate program, practicum
setting, weeks to date with current supervision group, and number of members in
current supervision group. Information was also gathered on supervisor gender, age,
ethnicity, and level of physical attractiveness and professional appearance.
50
Group Climate Questionnaire-Short Form. The GCQ (MacKenzie, 1981) is a
12-item instrument designed for completion at the end of a group session. Three
factored scales are reported: 1) Engaged (a positive working environment), 2)
Conflict (a negative atmosphere with anger and distrust), and 3) Avoiding (members
avoiding personal responsibility for group work). The items are scored on a 6-point
Likert-scale ranging from "not at all" to "always/definitely." The shortened version
is derived from the GCQ-L, and consists of items that were found to have the highest
factor loadings for these three factors (MacKenzie, 1981). MacKenzie (1981) stated
that the GCQ-L was reliable and that results in clinical settings support the construct
validity based upon item-scale correlations and comparison to therapists' ratings.
For the purposes of this study, only items on the "engagement" scale were used.
Items on the Engagement scale are related to cohesion, the member's desire to attend
the group, the importance of the group to them, and their sense of close intense
participation. The Cronbach alpha coefficient was .80 for this sample.
Supervision Satisfaction Questionnaire. The SSQ is an 8-item scale adapted by
Ladany, Hill, Corbett, and Nutt (1996) from the Client Satisfaction Questionnaire
(Larsen, Attkinsson, Hargreaves, & Nguyen, 1979). Participants appraise their
satisfaction with various aspects of supervision on a 6 point scale (e.g., 1= No.
definitely not, 6 =Yes. definitely). The SSQ has been found to be negatively related
to supervisee perception of frequency of supervisor ethical violations and inversely
related to supervisee nondisclosures involving negative reactions to the supervisor
(Ladany et al, 1996). The Cronbach alpha coefficient was .97 for this sample.
51
Supervisory Working Alliance Inventory-Client Focus subscale. The SWAI is
a 19-item scale created by Efstation et al. (1990) to measure the working alliance
between supervisee and supervisor. Trainee scores were reported to measure the
factors of "rapport" and "client focus." The S W AI was found to have adequate
convergent and divergent validity based on comparison to scales from the
Supervisory Styles Inventory (Friedlander & Ward, 1984). For the purposes of this
study, only the scale for "client focus" has been used, and will be referred to as the
"SWAI-CF." This subscale was reported to measure the emphasis the supervisors
placed on promoting the trainees' understanding of the client. The Cronbach alpha
coefficient was .90 for this sample.
Procedure: The investigator constructed an internet based survey using a
software package called "Survey Solutions for the Web" by Perseus Software (see
Appendix A). As noted previously, 219 training directors were contacted via
electronic mail and asked to forward an email to their students describing the
purpose of the study and the nature of their participation.
Students who were interested in completing the questionnaire were asked to
follow a hyperlink to the website. The website contained detailed information of the
study in accordance to the IRB requirements. Completion of the survey served as the
students' consent to participate in this research study.
To complete the survey, participants either typed in information, or checked
offboxes on the website. The format was similar to a traditional paper and pencil
survey. When the participant was finished, their participation was confirmed upon
52
submission of their survey. The data was then routed to a secure server, and then
sent via email directly to the investigator. No identifying information was collected
on the survey or electronically, ensuring confidentiality and anonymity. The
software program then imported the data from the electronic mail directly into a
spreadsheet compatible with SPSS.
Two hundred nineteen training directors were contacted through the Council of
Counseling Psychology Training Programs (CCPTP) and the Association of
Counseling Center Training Agencies (ACCT A) and asked whether they would be
willing to forward a survey via email to their eligible master's and doctoral trainees
who may be participating in group supervision. Respondents were considered
eligible if they were currently receiving group supervision. It is unknown how many
supervisees received this email. However, based on the number of students from
different training programs, it would appear that this Internet based survey was well
distributed across the United States. For purposes of confidentiality, the participants
were given the option to identify the university that they were attending. Of the 227
participants, 53 did not complete this item, and of the 174 who did fill it out, 64
university programs were represented.
53
Chapter III
Results
This chapter will outline the statistical procedures and results used to
determine the psychometric properties of the Group Supervision Scale. The results
of the exploratory factor analysis will be presented to show the underlying structure
of the GSS, and the results of several univariate multiple regressions will be
presented as evidence for concurrent validity.
Factor Analysis
Scores on the Group Supervision Scale were subjected to iterative principle
components extraction with the SPSS-X statistical computing package. The Kaiser
Meyer-Olkin measure of sampling adequacy (Kaiser, 1974) was .95 which indicates
that factor analysis was appropriate for the data set.
For the Group Supervision Scale the researcher considered several "stopping
rules" that are used in factor analysis as outlined by Bryant and Y arnold (1998).
Using the Kaiser-Guttman retention criterion, two factors were yielded with
eigenvalues greater than 1.0 and accounted for 73% of the variance. Cattell (1966)
proposed a graphical procedure, the scree test, for determining the appropriate
number of eigenvectors to extract in which the eigenvalues in the steep descent are
retained, and to drop all further components after the one starting the more gradual
descent. Results of Cattell's scree test conducted on the data set yielded three major
factors in the GSS.
54
According to Bryant and Yamold (1998), some researchers follow the
stopping rule using a predetermined percentage of variance explained (usually 75%).
Under this criterion, three factors accounting for 79% of the variance were extracted
from the GSS. Finally, under the Joliffe rule (1986), a more liberal criterion, a
researcher may drop all components with eigenvalues under .70. Using the Joliffe
rule, three components were kept from the data set.
Based on the factor loadings, and the high (10.53) initial eigenvalue of the
first component, the researcher considered retaining only one factor and defining the
GSS as a unidimensional scale. However, according to Pedhauzer and Schmelkin
(1991) and Bryant and Yamold (1988), the comprehensibility of the decision is the
most meaningful criterion. Though not a mathematical criterion, they suggest
determining the number of factors to those whose dimension of meaning is readily
comprehensible. A unidimensional or single factor model did not seem
comprehensible given the item pool of the GSS. Because the third component had
an initial eigenvalue close to one (.92), and the results based on the Cattell scree test,
the Joliffe rule, and the criterion of variance explained, the researcher decided to
retain three factors for the Group Supervision Scale.
The three factor solution were then subjected to orthogonal (varimax)
rotation. Note that the researcher also requested oblique rotations. However, the
oblique rotations did not result in a comprehensible factor model. fu addition,
orthogonal rotation was chosen because the interitem correlations were moderate to
high (Pedhauzer & Schmelkin, 1991).
55
One rule of thumb used by researchers is that factor loadings exceeding .4
and .5 as meaningful, however Norman and Streiner (1994) stated that instruments
using Likert-type scales a .6 might be required to be considered "high." As
mentioned previously, the factor analysis extracted three components. Several of the
items appeared to load moderately on the first component. Items 1, 3, and 12
appeared to load moderately (>.50) on Factor 1, however these items also loaded
highly (>.6) on the second or third factor. For the reasons noted above, the
researcher decided to group the three components based on items that rated higher
than .60 on the factor loadings. Item 13 did not load highly on any factor, but had its
highest loading (.558) on Factor 1.
Factor 1 of the Group Supervision Scale accounted for approximately 28% of
the known variance. The 4 items that loaded highly (>.60) on this factor reflected
the supervisor's attention to group safety, and was named Group Safety. Factor 2
accounted for approximately 26% of the known variance. The six items that loaded
highly (>.60) on this factor reflected the emphasis the supervisor placed developing
counseling intervention skills and the ability to conceptualize a case, and was named
Skill Development and Case Conceptualization. Factor 3 of the GSS accounted for
approximately 25% of the known variance. The 5 items that loaded highly (>.60) on
this factor reflects the supervisors attention to group management (e.g., time
management, structure, direction, focus), and was named Group Management. Table
1 presents the factor loadings, eigenvalues, and percentages of variance accounted
for by each factor as well as the item-scale correlations.
56
Table 1
Factor Loadings, Eigenvalues, Variance Explained, and Item-Scale Correlations in
thG S Sal e roup uperv1s1on c e
Factor 1 2 3 Item-
Item Scale
1. The supervisor provides useful feedback .56 .62 .30 .82
regarding my skills and interventions.
2. The supervisor provided helpful .43 .75 .30 .81
suggestions and information related to client
treatment.
3. The supervisor facilitates constructive .53 .62 .31 .81
exploration of ideas and techniques for
working with clients.
4. The supervisor provides helpful .10 .79 .40 .68
information regarding case
conceptualization and diagnosis.
5. The supervisor helps me comprehend and .38 .79 .29 .80
formulate clients' central issues.
6. The supervisor helps me to understand the .48 .70 .26 .80
thoughts, feelings, and behaviors of my
clients.
7. The supervisor encourages trainee self- .75 .30 .30 .76
exploration appropriately.
8. The supervisor enables me to express .82 .31 .15 .74
opinions, questions, and concerns about my
counseling.
9. The supervisor created a safe environment .78 .27 .36 .82
for group supervision.
10. The supervisor is attentive to group .74 .31 .33 .79
dynamics.
11. The supervisor effectively sets limits, .46 .34 .64 .81
and establishes norms and boundaries for
the group.
12. The supervisor provides helpful .54 .37 .64 .87
leadership for the group.
13. The supervisor encourages supervisees .56 .27 .46 .70
to provide each other feedback
14. The supervisor redirects the discussion .37 .29 .77 .78
when appropriate.
15. The supervisor manages time well .16 .27 .85 .65
between all the group members.
16. The supervisor provides enough .29 .32 .83 .77
structure in the group supervision.
57
I
Table 1 continued
Eigenvalue 10.53 1.12 .92 I I
Percentage of Variance 28.90 25.00 24.60
N=227 Bold numbers represent the highest factor loading for each item.
Note: Factor 1 = Group Safety, Factor 2= Skill Development and Case
Conceptualization, and Factor 3 = Group Management.
Preliminary Analyses
To test the potential confounding influence of demographic variables on the
primary variables of interest (i.e., scores on the SSQ, the "client focus" subscale of
the SWAI, GCQ, and the total score on the GSS), a series of multivariate regression
analyses were conducted. In each analysis, the demographic variable (i.e., race,
gender, or age) serves as the independent variable, and the variables of interest
served as the dependent variable. The results show that the participant's age, race,
gender, year in program, total semesters of counseling experience, and degree
objective were not significantly related to their total scores on the Group Supervision
Scale, E(6, 219) = .340, .Q = .92, r
2
= .01; the Group Climate Measure, f(6, 219) =
1.85, .Q = .09,? = .05; the Satisfaction with Supervision Questionnaire, E(6, 219) =
.90, .Q = .50, ? = .02; and the "client focus" subscale of the Supervisory Working
Alliance Inventory, E(6, 219) = .138, .Q = .99,? = .004.
Descriptive Statistics
The Group Supervision Scale is a self-report measure containing 16 items
rated on a 6-point Likert-type format (e.g., 1 =strongly disagree to 6 =strongly
agree). The scale consisted of three subscales with 4 items addressing Group Safety,
58
6 items addressing Skill Development and Case Conceptualization, and 6 items
addressing Group Management.
The mean scores for the three factors of the GSS were as follows: Group
Safety (M = 21.99, SD= 6.41), Skill Development and Case Conceptualization (M =
26.83, SD = 6.85), and Group Management (M = 21.50, SD = 6.50). For the overall
score of the Group Supervision Scale, the average score was 70.32 (SD = 18.27) for
this sample. Mean scores for the other instruments used for convergent validity were
also found for the Group Climate Questionnaire (M = 21.42, SD = 4.53), the client
focus subscale of the Supervisory Working Alliance Inventory (M = 28.69, SD =
7.73), and the Supervision Satisfaction Questionnaire (M = 35.12, SD = 10.46).
Histograms (see Appendix B) show that participants tended to rate their supervisors
favorably on these dimensions.
Cronbach's alpha was used to estimate the internal consistency reliability of
each of the three GSS factors. Alpha coefficients for the Group Supervision Scale
were .91 for Group Safety, .94 for Skill Development and Case Conceptualization,
and .93 for Group Management. The Alpha coefficient for the total GSS was .96 for
this study.
The item-scale correlations for the three factors were computed as well. For
the Group Safety factor, the correlations ranged from .78 to .83, from .73 to .87 for
Skill Development and Case Conceptualization, and from .77 to .88 for Group
Management. The correlations ranged from .68 to .88 for the total Group
Supervision Scale (see Table 1).
59
A preliminaty analysis of the relationship between the Group Supervision
Scale and the three scales (Group Climate Measure, Satisfaction with Supervision
Questionnaire, and the client focus scale of the Supervisory Working Alliance
Inventory) used to estimate convergent validity. High correlations were found
between the total score on the Group Supervision Scale and scores on the Group
Climate Questionnaire (I = . 72, n < . 01 ), the Satisfaction with Supervision
Questionnaire (I= .90, n < .01), and the client focus scale of the Supervisory
Working Alliance Inventory (I= .86, n < .01). Regarding divergent validity, the
Group Supervision Scale were not found to be significantly related to supervisor's
age (I= -.07, n < .01), supervisor's attractiveness (I= .08, n < .01), or the
supervisor's professional appearance (I= .05, n < .01).
Examination of the three factors of the Group Supervision Scale yielded
correlations as expected. Factor I (Group Safety) was highly correlated with the
Group Climate Questionnaire (I= . 725, n < .01 ). Factor 2 (Skill Development and
Case Conceptualization) was highly correlated with the "client focus" scale of the
SW AI (I= .87, n < .01). And Factor 3 (Group Management) was highly correlated
with the Satisfaction with Supervision Questionnaire (I =.82, n < .01). Means and
intercorrelations for the three factors of the Group Supervision Scale and the other
variables studied are shown in Table 2.
Table2
Means, Standard Deviations, and Intercorrelations
1 2 3 4 5 6 7 8 9 10
1 1
2 .93* 1
3 .93* .80* 1
60
Table 2 continued
4 .92* .70* .77* 1
5 .72* .73* .67* .60* 1
6 .90* .83* .85* .82* .70* 1
7 .86* .73* .87* .79* .62* .83* 1
8 -.07 -.19* -.08 -.10 -.08 -.06 -.02 1
9 .05 .04 .04 .06 .02 .05 .07 .04 1
10 .08 .11 .07 .03 -.01 .04 .05 -.28* .19* 1
M 70.32 21.99 26.83 21.50 21.42 35.12 28.69 43.61 4.62 3.82
so 18.27 6.41 6.84 6.50 4.53 10.46 7.73 8.48 1.12 .97
*p<.Ol
Note: 1 =Group Supervision Scale (GSS) Total, 2 = GSS Factor 1, 3 = GSS Factor
2, 4 = GSS Factor 3, 5 = Group Climate Questionnaire, 6 = Satisfaction with
Supervision Questionnaire, 1 = Supervisory Working Alliance Inventory-Client
Focus Subscale, 8 =Supervisor's Age, 9 =Supervisor's Appearance, and 10 =
Supervisor's Physical Attractiveness.
After inspecting the high intercorrelations in Table 2, the researcher entered
the primary variables into a simultaneous regression analysis. The high preliminary
correlations suggested that the different instruments used in this study were all
measuring a similar latent construct. A regression analysis using standardized scores
was conducted to control for the potential intercorrelations among the predictor
variables (GCQ, SSQ, and SWAI) and the criterion variable (GSS), and to determine
the partial correlation coefficient (rp) of each instrument. That is, the researcher was
interested in finding the unique contribution of each instrument when all of the other
instruments in the study have been partialed out. Each of these instruments was
theorized to measure a different construct (e.g. group cohesion, general satisfaction,
and task orientation).
The multiple correlation coefficient between the three predictor variables
(GCQ, SSQ, and SWAI) and total GSS scores was .93, .E(3, 222) = 486.64, n <.01,
61
which accounted for approximately 87% of the variance. The Group Supervision
Scale was significantly related to the Group Climate Questionnaire(!= 4.29,12 <.01),
the Satisfaction with Supervision Questionnaire(!= 10.65,12 <.01), and the "client
focus" scale of the SW AI (! = 8. 06. 12 < . 0 I). The partial correlation coefficients for
each instrument related to the Group Supervision Scale were as follows: Group
Climate Questionnaire (rp = .28), the "client focus" scale of the SWAI (rp = .48), and
the Satisfaction with Supervision Questionnaire (rp =.58).
In addition. the researcher requested collinearity diagnostics. According to
Tabachnick and Fidell (1996), criteria for multicollinearity are a conditioning index
> 30 and at least two variance proportions >.50 for a given variable. In this
regression analysis, the conditioning indexes for the GCQ, SW AI-CF, and the SSQ
were 1.56, 2.45, and 3.86 respectively, and only one of the variance proportions were
greater than .5 (.00, 1.00, .00, and .00 respectively), indicating that multicollinearity
was within tolerance levels.
Construct Validity
Univariate multiple regression analyses were used to test the primary
hypotheses of this study. Univariate multiple regression analyses were used for all
of the major analyses to control for the potential intercorrelations among the
predictor variable and among the criterion variables (Haase & Ellis, 1987). Results
of these analyses are discussed in the following sections (see Table 3). Both
simultaneous and stepwise regression analyses were conducted to find the unique
62
contributions of each factor, and to detennine the best predictor of each criterion
variable.
Table 3
Results of the Univariate Simultaneous Regression Analyses: 3 Factors of the GSS
and Group Cohesion, Skill Development, and Satisfaction with Supervision.
ariables t
roup Climate Questionnaire (GCQ)
actor 1 (Group Safety) .43 5.43 <.01*
actor 2 (Skill Development) .27 3.21 <.01*
actor 3 (Group Management) .07 .94 .35
AI-Ciient Focus subscale
actor 1 (Group Safety) -.05 -.95 .34
actor 2 (Skill Development) .67 11.94 <.01*
actor 3 (Group Management) .31 6.11 <.01*
atisfaction with Supervision Questionnaire
SSQ)
actor 1 (Group Safety) .31 6.35 <.01*
actor 2 (Skill Development) .37 7.10 <.01*
.30 6.30 <.01*
Note: Factors 1-3 refer to the factors of the Group Supervision Scale and were used
as the predictor variables. The GCQ, SW AI, and SSQ were entered as the criterion
variable in each analysis. Bold numbers represent the highest standardized partial
regression weights in each analysis.
Group Supervision Scale and Group Cohesion
Two separate univariate multiple regression analyses were used to test the
hypothesized relationship between the three factors of the Group Supervision Scale
and the Group Climate Questionnaire. In each analysis the predictor variables were
the scores on the group safety, skill development/case conceptualization, and the
63
group management subscales of the GSS, and the criterion variable were the scores
on the Group Climate Measure. The researcher first conducted a simultaneous
regression to determine the unique contributions of each factor of the GSS to group
cohesion, and a stepwise regression was used to determine which factor would be the
best predictor of group cohesion.
For the simultaneous regression, the overall proportion of variance of the
group supervisor's influence on group cohesion accounted for by these factors was
significant, 1:(3, 222) = 83.13, R < .01, fl =.55. The standardized partial regression
weight(~) of Factor 1 (Group Safety) was .43 and significant(!= 5.43, R < .01).
Factor 2 (Skill development/case conceptualization) was also significant(~= .27, 1 =
3.21, n < .01). Factor 3 (Group Management) was not significantly related to group
cohesion(~= .07, 1 = .94, n = .35).
Using a stepwise regression, only one variable was entered into the prediction
equation, Factor 1 (Group Safety). The overall proportion of variance accounted for
by Factor 1 was significant, E (1, 224) = 217.08, n < .01, r = .49. Using this single
predictor, 49% of the variance in group cohesion was accounted for. Using Factor 1
and 2 as predictors, 52% of the variance in group cohesion was accounted for (an
increase of3%). Therefore, Factor 1 (Group Safety) was found to be the best
predictor of group cohesion.
Groun Su,PeiYision Scale and skill/case conceytualization develomnent
Two univariate multiple regression analysis were used to test the
hypothesized relationship between the three factors of the Group Supervision Scale
64
and the "client focus" scale of the SW AI. The predictor variables were the scores on
the group safety, skill development/case conceptualization, and the group
management subscales of the GSS, and the criterion variable were the scores on the
"client focus" scale of the SW AI. A simultaneous regression was first conducted to
determine the unique contributions of each factor of the GSS to skill development,
and a stepwise regression was used to determine which factor would be the best
predictor of skill development.
In the simultaneous regression the overall proportion of variance of the group
supervisor's ratings on "client focus" accounted for by these three factors was
significant, .E(3, 222) = 281.91,12 < .01, r
2
=.79. The standardized partial regression
weight (J3) of Factor 1 (Group Safety) was -.05 and not significant(!= -.95, 12 = .34).
Factor 2 (Skill development/case conceptualization) was highly significant (J3 = .67, !
= 11.95,12 < .01). Factor 3 (Group Management) was also significantly related to
"client focus" (J3 = .31,! = 6.11, 12 < .01).
The stepwise regression entered only one variable into the prediction
equation, Factor 2 (Skill Development and Case Conceptualization). The overall
proportion of variance predicted by Factor 2 was significant, .E(l, 224) = 692.74,12 <
.01, r.
2
= .76. Using Factor 2 as a predictor, 75% of the variance in skill development
as measured by the SW AI-CF was accounted for, compared to 79% accounted for by
Factor 2 and Factor 3 (Group Management) as predictors. Therefore, Factor 2 (Skill
Development and Case Conceptualization) was found to be the best predictor of skill
development as measured by the "client focus" subscale of the SW AI.
65
Group Supervision Scale and the satisfaction with supervision
Two univariate multiple regression analysis was used to test the hypothesized
relationship between the three factors of the Group Supervision Scale and
satisfaction with supervision. The predictor variables were the scores on the group
safety, skill development/case conceptualization, and the group management
subscales of the GSS, and the criterion variable were the scores on the Satisfaction
with Supervision Questionnaire. A simultaneous regression was first conducted to
determine the unique contributions of each factor of the GSS to overall satisfaction,
and a stepwise regression was used to determine which factor would be the best
predictor of satisfaction.
In the simultaneous regression the overall proportion of variance of the group
supervisor's influence on supervision satisfaction accounted for by these factors was
significant, !:(3, 222) = 337.06, Q < .01, r =.82. The standardized partial regression
weight (f3) of Factor 1 (Group Safety) was .31 and significant (1 = 6.35, Q < .01).
Factor 2 (Skill development/case conceptualization) was also significant (f3 = .37,! =
7.1 0, Q < .01 ). Factor 3 (Group Management) was also found to be significantly
related to supervision satisfaction (f3 = .30, t = 6.30, Q < .01).
The stepwise regression entered only one variable into the prediction
equation, Factor 2 (Skill Development and Case Conceptualization). The overall
proportion of variance predicted by Factor 2 was significant, f(l, 224) = 578.49, Q <
.01, ( = .72. Using Factor 2 as a predictor, 72% of the variance in satisfaction was
66
accounted for, compared to 78% accounted for by Factor 1 and Factor 2 as
predictors, and 82% by all three factors.
Group Supervision Scale and Attractiveness. Appearance. and Age
Table 1 lists the intercorrelations between the Group Supervision Scale and
the participants' ratings of their group supervisors in physical attractiveness (r = .08),
professional appearance (r = .05), and age (r = -.07). Physical attractiveness,
professional appearance, and age were not found to be significantly related to scores
ontheGSS.
In summary, there were four validity hypotheses tested in this investigation.
They were: (1) There would be a significant, positive relationship between the
perfonnance of the group supervisor and the supervisee's perception of group
cohesion; (2) There would be a significant. positive relationship between the
perfonnance of the group supervisor and the supervisee's perception of the
supervisor's impact on skill development and case conceptualization; (3) There
would be a significant, positive relationship between the perfonnance of the group
supervisor and the supervisee's overall sense of satisfaction with supervision; and ( 4)
There would be no significant correlation between the perfonnance of the group
supervisor, the age of the supervisor, and ratings of the group supervisor's physical
attractiveness and professional appearance. The results indicate that all four
hypotheses were supported.
67
Interaction Effects
The researcher tested for interaction effects using a two way ANOV A as
outlined by Stockburger (1997) and Tabachnick and Fidell (1996). The raw scores
of Factors 1, 2, and 3 were recoded to reflect high and low scores (by median splits)
and then analyzed using a GLM general factorial model for regression analysis and
analysis of variance for one dependent variable by the three factors. In each
analysis, the instruments used for convergent validity (GCQ, SW AI-CF, and SSQ)
were entered as the dependent variable, and the three factors of the Group
Supervision Scale (Group Safety, Skill Development & Case Conceptualization, and
Group Management) were entered as the fixed factors. All three analyses revealed
no significant interaction effects between the three factors of the GSS and the other
instruments. The results of this analysis can be seen on Table 4.
Table 4
Two Way ANOV A Tests for Main Effects and Interactions
Type III Sum of df Mean Square F Sig.
Source Squares
Group Climate Questionnaire
GSS.lHL 282.42 1.00 282.42 22.01 0.00
GSS.2HL 114.19 1.00 114.19 8.90 0.00
GSS.3.HL 8.24 1.00 8.24 0.64 0.42
GSS.1HL* 0.49 1.00 0.49 0.04 0.85
GSS.2HL
GSS.lHL * 8.36 1.00 8.36 0.65 0.42
GSS.3.HL
GSS.2HL* 5.63 1.00 5.63 0.44 0.51
GSS.3.HL
68
Table 4 continued
Supervisory Working Alliance-
Client Focus Subscale
GSS.1HL 45.17 1.00 45.17 1.56 0.21
GSS.2HL 1287.59 1.00 1287.59 44.39 0.00
GSS.3.HL 326.73 1.00 326.73 11.26 0.00
GSS.1HL * 10.65 1.00 10.65 0.37 0.55
GSS.2HL
GSS.1HL * 46.19 1.00 46.19 1.59 0.21
GSS.3.HL
GSS.2HL* 4.93 1.00 4.93 0.17 0.68
GSS.3.HL
Supervision Satisfaction
Questionnaire
GSS.1HL 770.43 1.00 770.43 17.02 0.00
GSS.2HL 1755.78 1.00 1755.78 38.79 0.00
GSS.3.HL 1020.55 1.00 1020.55 22.55 0.00
GSS.1HL * 22.98 1.00 22.98 0.51 0.48
GSS.2HL
GSS.1HL * 49.68 1.00 49.68 1.10 0.30
GSS.3.HL
GSS.2HL* 2.58 1.00 2.58 0.06 0.81
GSS.3.HL
Note: For Main Effects, GSS.1HL =Group Supervision Scale Factor 1 High/Low,
GSS.2HL = Group Supervision Scale Factor 2 High/Low, and GSS.3HL = Group
Supervision Scale Factor 3 High/Low. * denotes interaction effects between the
above factors.
69
ChapterN
Discussion
The role of the group supervisor is similar to but different than that of an
individual supervisor. Both Counselman and Gumpert (1993) and Hayes et al.
(2001) noted that effective group supervision is not simply individual supervision
within the presence of a group. An effective group supervisor provides the
ingredients of good individual supervision, and combines this with the skills of a
good group leader. Bernard's (1979) Discrimination Model for individual
supervision proposes that there are three main functions of supervision: to develop
the counselor trainee's process, conceptualization, and personalization skills.
Regarding group supervision, Holloway and Johnston (1985) found that most
supervision groups involved these three domains of focus: 1) didactic, 2) case
conceptualization, and 3) interpersonal process. These domains can be seen as
similar to those outlined by Bernard (1979), with the added dimension of the
interpersonal dynamics that are inherent in group formats.
The literature on group supervision indicates that group supervision is not
simply individual supervision in the presence of a group. Wilbur et al. (1991) and
Counselman and Gumpert (1993) noted that when group supervision is used to
address individual concerns one at a time, the unique characteristics of the group
format are not utilized. An effective group supervisor combines the necessary
components of individual supervision (conceptualization and skill development,
providing didactic information, and personal growth) with those skills required by a
70
group leader to facilitate group dynamics and promote group cohesion. The effective
group supervisor promotes a supervisory experience that can be potentially richer
than individual supervision. Group supervision, when used effectively, can provide
a wider array of perspectives and strategies, broaden theoretical understandings,
teach group process through experiential learning. provide vicarious learning
opportunities, and personal growth (Borders, 1991 ~ Counselman & Gumpert, 1993 ~
Hayes et al. 2001 ). For this reason, Kruger et al. (1988) argued that group
supervision may be a more complex task than individual supervision.
Overall, this study indicates that supervisees are satisfied with their group
supervision, gain valuable help with their intervention skills and ability to understand
a client, and that they believe that their group supervisors facilitate cohesion in their
supervision groups. In addition, the results indicate that overall satisfaction with
supervision is related to both the sense of group cohesion and the gaining ofvaluable
clinical skills. These findings support those found in other studies related to the
efficacy of group supervision (Savikas et al., 1986~ Kruger et al .• 1988~ Ray &
Altekruse, 2000), prior studies investigating group cohesion (Emerson, 1996; Walter
& Young. 1999), and the general belief among clinicians that group supervision has
a great deal to offer (Bernard & Goodyear, 1998). In light of these findings, the
ability to evaluate a group supervisor along the dimensions of group management,
development of group safety, and the development of intervention and
conceptualization skills may yield important information for supervisors and trainees
engaged in ongoing group supervision.
71
This study created an instrument that measured the performance of the group
supervisor. With regard to the psychometric properties of the Group Supervision
Scale, the results indicate evidence of reliability and validity for the measure.
Evidence of the GSS' validity was supported by exploratory factor analysis and by
significant associations with theoretically relevant variables. The alpha coefficient
of the Group Supervision Scale was high (r = .96) indicating a high internal
consistency. The three factors (group safety, skill and case conceptualization
development, and group management) accounted for 79% of the variance.
While several of the individual items appeared to have moderate factor
loadings on the first component, results of the univariate multiple regression analyses
indicate that the different factors as described under the three factor solution are
conceptually distinct, and the use of "stopping" criteria outlined in the results section
also support a three factor model. The standardized partial regression weights (j3)
indicate the individual contributions of each factor when controlling for the effects of
the other factors. The univariate multiple regression analyses showed that for group
cohesion, as measured by the Group Climate Questionnaire (GCQ), Factor 1 (group
safety) had the highest standardized partial regression weight (j3). Likewise, when
examining the relationship between the three factors of the GSS and the "client
focus" subscale of the SW AI, the univariate multiple regression analysis showed that
Factor 2 (skill and conceptualization development) had the highest standardized
partial correlation weight (j3). In addition, the regression analyses showed that
Factor 1 (group safety) was not significantly related to the clinical focus as measured
72
by the "client focus" subscale of the SWAI, and that Factor 3 (group management)
was not significantly related to group cohesion.
The findings of the simultaneous regression analyses support the conceptual
distinctiveness of the three factors. The items found in Factor I (group safety) appear
to be related to creating a safe group environment and promoting personal growth,
which matches the findings that this factor would be most related to group cohesion,
and not significantly related to a focus on conceptualization and skill development.
Factor 2 (skill/case conceptualization development) had the highest partial regression
weight in relation to the "client focus" subscale of the SW AI, which reportedly
measures the emphasis the supervisor places on the development of clinical skills
and conceptualization ability. The items that comprise Factor 3 (group management)
appear to reflect time management and structure imposed by the group leader, which
would not be expected to be highly related to group cohesion or clinical skills
development.
Additionally, the results of the stepwise regression analyses support the
distinctiveness of the three factors. In these analyses, the three factors were entered
into a stepwise regression analysis as the predictor variables, and the scores on the
GCQ, SWAI-CF, and SSQ were entered as the criterion variables. For group
cohesion, as measured by the GCQ, the best predictor was found to be Factor I
(Group Safety). Likewise, regarding skill development, as measured by the SWAI
CF, the best predictor was found to be Factor 2 (Skill Development and Case
Conceptualization).
73
The three factors of the Group Supervision Scale also appear to be supported
by the group supervision literature. Sansbury (1982) suggested four group
supervision tasks: 1) teaching interventions directed at the entire group (didactic), 2)
presenting specific case-oriented information, suggestions and feedback (case
presentation), 3) focusing on affective responses of the supervisee as pertains to the
client (personal growth), and 4) processing the group's interaction and development
(group facilitation). Similarly, Holloway and Johnston (1985) found that most
supervision groups involved these three domains of focus: 1) didactic, 2) case
conceptualization, and 3) interpersonal process.
Regarding the GSS, the first factor (group safety) is most related to the
creation of cohesion and safety in the group, which promotes the process of personal
growth in the supervisee. The second factor (skill/conceptualization development)
relates to the didactic and case-presentation components of group supervision. And
the third factor (group management) relates to the facilitation of the group dynamics
and the management of such factors as time, structure, and the active direction of the
group.
Group Supervision Scale and Group Cohesion
The results supported the first hypothesis that effective group supervision is
related to stronger group cohesion. That is, that supervisors who effectively promote
group supervision are also evaluated highly as group supervisors. The idea that good
group supervisors are better able to promote group cohesion makes sense in light of
the findings that once supervisees are able to achieve more cohesion, they are better
74
able to utilize their group supervision experience to its advantages (Walter & Young,
1999). Similarly, Emerson ( 1996) stated that creating an atmosphere of safety and
trust allows supervisees to take risks in supervision and gain the feedback necessary
to lean from themselves and the other group members.
According to Prieto (1996), the interactions among the supervisees seemed to
be the primary and most influential focus of group supervision. Greater trust
encourages cooperation and increases productivity and achievement in groups
(Westwood, 1989). Bernard and Goodyear (1998) warn that two types of
interpersonal processes may occur during group supervision: competition or support
between supervisees.
Group supervisors should attempt to create group cohesion early in the
group's development. This leads to a greater sense of safety and trust, and allows the
supervisees to challenge themselves and challenge each other to gain the most out of
their group supervision experience. As a way to foster cohesion, the literature
suggests that the group supervisor first provide structure in the form of simple
exercises (Yalom, 1985; Caple & Cox, 1989). This initial structure helps improve
communication and reduces the ambiguity and anxiety felt by the supervisees in the
beginning stages of the group's development.
Based on those items of the GSS that were found to contribute to a sense of
group safety, the supervisor should be attentive to group dynamics; enable the
supervisees to express their opinions, questions, and concerns about their counseling;
and encourage self-exploration appropriately.
75
Group Supervision Scale and Satisfaction with Supervision
The results supported the second hypothesis that the supervisee's perception
of the group supervisor would be related to greater trainee satisfaction with
supervision. More specifically, the supervisee's perception of his or her group
supervisor's ability to create a safe group environment, enhance and develop their
counseling and conceptualization skills, and manage the supervision group were all
positively related to their overall satisfaction with supervision. That is, all three
components of the GSS were found to be significantly related to satisfaction with
supervision.
This finding is supported by the number of studies on group supervision that
suggest that the role of the group supervisor includes those responsibilities of
effective group leadership (Hayes et al., 2001), creating a safe group environment
(Counselman & Gumpert, 1993), and diagnostic expertise (McAuliffe, 1992). A
deficiency in any of these three areas can drastically affect the group supervision
experience. For example, in a supervision group that is not safe and cohesive, the
supervisees will not challenge themselves or each other, and will not provide each
other with constructive and helpful feedback. Or, a group that has developed a sense
of trust or cohesiveness, will not be helpful to the supervisee if the supervision does
not provide enhancement of clinical and conceptualization skills.
Group Supervision Scale and Skill and Case Conceptualization Development
The results also supported the third hypothesis that effective group
supervision practices are associated with perceptions that the group was effective in
76
skill and case conceptualization development. This shows that effective group
supervision is not solely comprised of the creation of safety and trust in the group.
Effective group supervision also promotes the development of counseling
intervention skills and the ability to understand or conceptualize clients. Watkins
(1997) stated that clinical supervision (individual or group) is important because it
provides supervisees with feedback about their performance, and gives them
alternate views and perspectives about patient dynamics, interventions, and courses
of treatment. Similarly, Wampold and Halloway (1997) believe that one of the main
goals of supervision is to increase the skill level of the supervisee.
However, group supervision is different from individual supervision in that
the potential for a broader range of observations, thoughts, and strategies may be
developed by the collective whole of the group, in comparison to individual
supervision (Counselman & Gumpert, 1993). A supervision group can provide a
greater array of theoretical perspectives, and can expose the supervisee to a multitude
of case formulations and treatment interventions. The advantages of group
supervision, however, cannot be utilized when group supervision is approached as
individual supervision in the presence of a group. The supervisor must effectively
lead the group and facilitate the interactions between the group members. Therefore
the management of the group via leadership and direction is an important function of
the group supervisor.
The univariate analyses of the factors comprising the Group Supervision
Scale showed that both the emphasis on client interventions and understanding, and
77
the management of the group were important factors related to the development of
the skill level of the supervisee. Based on those items that comprised these two
factors, the group supervisor should provide useful feedback, suggestions, and
information; encourage the supervisees to provide feedback to each other, and
provide helpful leadership and structure to the group.
Group Supervision Scale and Age. Attractiveness. and Appearance
The results supported the fourth hypothesis that effective group supervision
had no significant relationship to the group supervisor's age, physical attractiveness,
or professional appearance. Thus, effective group supervision can be seen as
independent of the supervisor's level of physical attractiveness, professional
appearance (e.g. how the supervisor dresses), or age. This finding is not surprising
given that these variables should not affect a supervisor's ability to provide effective
group supervision.
Heppner and Roehlke (1984) believed that supervisee's perceptions of
supervisor expertness, attractiveness, and trustworthiness change as the supervisee
becomes more sophisticated and is able to differentiate between what is being said,
and who is saying it. While the results in this study indicated that there were no
significant correlations between effective group supervision and the supervisor's age,
physical attractiveness, and professional appearance, there were some interesting
findings. For example, there was a significant, negative relationship between the
supervisor's age and level of physical attractiveness.
78
Implications for Theocy. Research. and Practice
In a national survey of predoctoral counseling psychology internship
programs, Rica and Cornish (1995) found that 65% of the sites conducted group
supervision. As mentioned several times in this study, group supervision has been
widely used, but poorly understood. In both Holloway and Johnston's (1985) and
Prieto's ( 1996) reviews of the group supervision literature, they found that the
majority of the studies used "author devised" questionnaires that had not been either
normed or validated. In addition, Bernard and Goodyear ( 1998) noted that almost all
of the instruments that have been used to assess supervision variables were originally
developed for other purposes, such as psychotherapy research, and that this practice
may change the psychometric properties of these instruments.
Prieto (1996) and Holmes et al. (1998) both concluded that investigators need
to develop and use supervision-specific assessment instruments that can enable the
analysis of the variablesinvolved in group supervision. At the time of this study,
there was only one prior study that attempted to evaluate the psychometric properties
of a scale that assessed aspects of group supervision (Calsyn et al., 1999). This
instrument consisted of four items, and had limited utility.
This study attempted to add to the small body of knowledge related to group
supervision, and to develop and provide preliminary validation evidence for an
instrument that evaluates the performance of the group supervisor. Analyses of the
Group Supervision Scale revealed three orthogonal factors, an the GSS scores
79
proved to have acceptable estimates of internal reliability and concurrent and
discriminant validity.
The validation of the GSS has implications for theoretical approaches to
group supervision. Effective group supervision entails the creation of a safe,
cohesive group; the enhancement and development of intervention skills and
conceptualization ability; and the facilitation and management of group dynamics.
Specific training for group supervision may be necessary. Most supervision training
focuses on those aspects that are common to individual supervision. The added
dimension of the group context requires additional training to utilize the full
potential of group supervision.
With this instrument, researchers can seek ways to further understand the
complex dynamics involved in group supervision. Given the general lack of
empirical studies on group supervision, there are several possibilities. For example,
a researcher can examine how differences in race, gender, developmental level, etc.
may affect the perceptions that the supervisee may have on the supervisor. In this
study, these variable did not appear to have a significant impact on the evaluation of
the supervisor. However, a more purposeful study would need to be conducted
examining these possible group differences.
Supervisor's scores on the GSS can be examined to determine whether they
vary systematically by the theoretical orientation of the participants and the
supervisors. This also suggests some implications for practice. Supervisors'
theoretical orientations may lead them to develop and stress different aspects of
80
group supervision. For example, a more process oriented supervisor may be more
attentive to group cohesion and group dynamics than to skills and conceptualization
training or feedback. It is important to stress that the various dimensions of group
supervision are important to the overall experience of group supervision.
Finally, it is important to note that the supervisor is just one component of the
group supervision experience. Future research could examine the supervisors impact
on the supervision group, as well as other possible variables that may be out of the
supervisor's control. For example, the supervisor may be evaluated highly on the
three components of the GSS, but the negative influence of one of the group
members may have a greater impact that the supervisor's ability or attempts to lead
the group.
Practical implications of the Group Supervision Scale include its use in
evaluation of the group supervisor. From this author's experience, evaluation of the
group supervisor was often conducted with the use of instruments that were
developed for individual supervision. These instruments did not address the
dimensions related to group dynamics or group management, which is an important
aspect of the group supervision process. Those instruments that have been
developed for individual supervision may be able to evaluate a group supervisor's
ability to provide good feedback and suggestions related to the supervisee's clinical
work, or the supervisee's perception of their working alliance or style, but these
instruments do hot evahiate the supervisor's role as group facilitator as well.
81
The GSS can be used for formative and summative evaluation of the group
supervisor. It can be administered early in the supervision group's lifespan, to
provide feedback to the supervisor on areas that may need to be addressed or
improved, or the GSS can be administered as a final evaluation at the completion of
the group.
The GSS can be used in the training of supervisors. As noted before, new
supervisors can benefit from discussion of these dimensions and recommendations
regarding how to facilitate their supervision groups to maximize their potential. In
particular, supervisors-in-training can use the GSS to assess their performance in
creating group cohesion/safety, managing the group dynamics, and developing their
supervisees' intervention skills and conceptual abilities. In the same way, more
experienced supervisors can use the GSS to determine those areas that need to be
emphasized when their supervision group does not appear to be developing as
expected.
Limitations of the Study
This study represents the preliminary attempts to develop and validate the
Group Supervision Scale. The stability of the three factor model of the GSS as well
as their relationship to the instruments used for validity purposes need to be
examined with a new sample of supervisees in both exploratory and confirmatory
factor analyses. Ultimately, the criterion for determining the number of factors
- 'sliou1d be the ieplicahility of the sobition. It IS important to extract only factors that
can be expected to replicate themselves when a new sample of subjects is employed.
82
Given the results of this study, a logical next step would be to edit the existing items,
and create additional items for each factor. Then, a confmnatory factor analysis
could be used to determine whether this factor structure conforms to what is
expected on the basis of the pre-establish theoretical foundation.
The high intercorrelations between the Group Supervision Questionnaire and
the SSQ, the SW AI-CF, and the GCQ were problematic. Multicollinearity may
cause statistical problems in a regression analysis, and creates greater difficulty in
portioning out the individual effects of the different variables. While the collinearity
diagnostics indicated that this was within tolerance levels, there may still cause
logical and statistical problems. These high intercorrelations may be a result of the
high overall ratings given to the supervisors across the different instruments (see
Appendix B for histograms). A larger sample size may be able to create more range
and variance in the results. In future studies, the use of different instruments for
convergent validity may also be more helpful.
Regarding the GSS in comparison to the other instruments, the high
intercorrelations also suggests that the GSS measures the same constructs as the
SW AI, SSQ, and GCQ. That is, that the GSS does not in fact evaluate the
"performance of the group supervisor," and instead measures the constructs of skill
development/case conceptualization, satisfaction with supervision, and group
cohesion. In this sense, the Group Supervision Questionnaire may not have the
utility that it was designed to measure.
83
In addition to that, the high intercorrelations do not support the uniqueness of
the Group Supervision Questionnaire in comparison to the other instruments. The
correlations suggest that the Group Supervision Questionnaire is, in fact, measuring
the same latent constructs that are being measured by the other instruments. While
this may be the case, it is important to note the content validity (Crocker & Algina,
1986) of the Group Supervision Scale. The items that comprise the GSS were
developed through a multi-step process. Focus groups were conducted to develop
items specific to the different aspects of group supervision. In addition, expert raters
were used to further support the content validity of those items. Therefore, while the
high intercorrelations between the GSS and the SSQ, SW AI, and GCQ suggest that
these instruments are all measuring the same construct, it is important to note how
the method of construction of the GSS also contributes to its uniqueness.
Another limitation of this study is that the trainee sample in this study mostly
consisted of white women and counseling psychology students. A more diverse
sample of racial groups, gender, and types of academic program would be needed to
examine the appropriateness of the GSS.
In addition, according to the group therapy literature (Tuckman & Jensen,
1971; Wheeler, 1997; Y alom, 1995), groups typically develop across a pattern of
predictable stages- (forming, storming, norming, etc.), and one may expect that there
would be changes over time in the groups development. Likewise, the supervisees'
perception of the group leader is dynamic, and would change as the group develops
over time.
84
A theoretical limitation of this study is the incorporation of theory taken from
the group counseling and therapy literature regarding the roles, responsibilities, and
skills of the group supervisor. Riva and Cornish's (1995) had made the suggestion
that "group counseling research can help guide research questions in group
supervision" (p. 524). However, Prieto (1997), in response to this suggestion, stated
that therapy-based perspectives may have the potential to be detrimental to research
in group supervision. Prieto (1997) reasserted the findings of Holloway and
Johnston (1985) that therapy-based group supervision environments or processes are
not effective approaches to use with trainees. This is due to the ethical concerns
related to a focus on the interpersonal process approach and the idea of "captive
therapy." In addition, Holloway and Johnston (1985) had concluded that there was
no empirical evidence that the interpersonal process approach to group supervision
increased trainee counseling skills.
Prieto ( 1997) also stated that the use of therapeutic constructs in
conceptualizing group supervision processes will cause epistemological confusion
that will result in inaccuracies in the knowledge base. For example, the use of
psychotherapy based instruments as dependent measures in group supervision
research can hinder the discovery and assessment on the unique constructs inherent
within group supervision. Prieto (1997) advocated the use of pedagogical constructs
and processes in conceptualizing group supervision since group supervision appears
to have a didactic and educational purpose, rather than a therapeutic goal. However,
85
in her critique, Prieto ( 1997) did not provide examples of these pedagogical
constructs.
However, Westwood (1989) took a different viewpoint and asserted that the
key components of group theory and practice should be used to help guide group
supervision. Westwood (1989) stated, "In establishing effective groups for
supervision it is recommended that specific aspects from the theory of groups (group
processes, leadership and stage development) be adopted by the instructor" (p. 351).
Ellis and Douce (1994) state that group supervisors "are encouraged not only to be
cognizant of the group literature but also to use group factors to optimize supervisor
learning" (p. 522). likewise, Wilbur et al. (1991) and Counselman and Gumpert
(1993) believed that group supervision can benefit from the use of processes found in
the literature for problem solving, personal growth, and process groups since group
supervision contains elements of these different types of groups.
Methodologically, a significant limitation of this study is due to the lack of
strong, empirically based measures of group supervision for which to compare this
instrument to and to establish concurrent and discriminant validity. No other known
instruments measuring any aspect of group supervision have been published,
normed, or validated. Therefore, this study was forced to compare this instrument to
those found in the group counseling domain. The Group Climate Questionnaire, the
Satisfaction with Supervision Questionnaire, and the Supervisory Working Alliance
Inventory were originally developed to measure constructs that were related to
86
group counseling and group therapy. The wording of the items and the intent of the
instruments do not reflect the specific issues related to group supervision.
In addition to this, the author has chosen to gather data from individual
participants rather than intact supervision groups. This may cause problems in that
different individuals may be rating the same supervisors. There may be different
scores related to the group climate or the group supervisor, even though these ratings
were made on the same groups and supervisors. According to MacKenzie (1981),
each member has his or her own perception of the group. "For that member, his
view is the group reality, even though others may see it differently" (MacKenzie,
1981, p287). A person who is poorly incorporated into the group, such as an isolate
or a scapegoat, may rate the group differently than the way others may rate a similar
dimension, such as cohesion or supportiveness.
As noted previously, the overall purpose of this study was to develop and
provide preliminary validation evidence of the Group Supervision Scale. Future
researchers would do well to account for these limitations and expand on these
findings.
87
References
Agnew, T., Vaught, C.C., Getz, H. G., & Fortune, J. (2000). Peer group clinical
supervision program fosters confidence and professionalism. Professional School
Counselin&, 4(1) 6-12.
Altfeld, D. A. (1999). An experiential group model for psychotherapy
supervision. International Journal of Group Psychotherapy, 49(2) 237-254.
Averitt, J. (1989). Individual versus group supervision of counselor trainees.
Dissertation Abstracts International, 50, 624.
Barrow, M., Domingo, R. A. (1997). The effectiveness of training clinical
supervisors in conducting the supervisory conference. Clinical Supervisor, 16(1) 55-
78.
Bernard, J.M. (1997), The discrimination model. In C.E. Watkins' (Ed.)
Handbook of psychotherapy supervision. New York: John Wiley & Sons.
Bernard, J.M. & Goodyear, R.K. (1998). Fundamentals of clinical supervision
(2nd Edition). Needham Heights, MA: Allyn & Bacon.
Blocher, D.H. (1983). Toward a cognitive developmental approach to
counseling supervision. Counselin& Psychologist, 11, 27-34.
Borders, L. D. (1991). A systematic approach to peer group supervision.
Journal of Counselin& & Development, 69(3) 248-252.
Bordin, E.S. (1983). A working alliance based model of supervision. The
Counselin& Psychologist, 11, 35-41.
Bryant, F.B. & Yaronld, P.R. (1998). Prinicple components analysis and
exploratory and confirmatory factor analysis. In Grim & Yamold's (Eds.) Reading
and understanding multivariate statistics. Washington D.C.: American
Psychological Association.
Calsyn, R. J., Burger, G. K., & Winter, J.P. (1999). Evaluating team leaders
and peers in group supervision. Clinical Supervisor, 18(2) 203-210.
Caple, R. B. & Cox, P. L. (1989). Relationships among group structure,
member expectations, attraction to group, and satisfaction with the group experience.
Journal for Specialists in Group Work, 14(1) 16-24.
88
Christensen, T.M. & Kline, W.B. (2000). A qualitative investigation of the
process of group supervision with group counselors. Journal for Specialists in Group
Work, 25(4), 376-393.
Cooper, L. & Gustafson, J.P. (1985). Supervision in a group: An application of
group theory. The Clinical Supervisor, 3(2), 7-25.
Counselman, E. F. & Gumpert, P. (1993). Psychotherapy supervision in small
leader-led groups. Group. 1993 Spr Vol17(1) 25-32.
Crocker, L.M. & Algina, J. (1986). Introduction to classical and modern test
theory. Orlando, FLA: Harcourt Brace Jovanocich, Inc.
Crutchfield, L. B., Price, C. B., McGarity, D., Pennington, D., Richardson, J.,
& Tsolis, A. (1997). Challenge and support: Group supervision for school
counselors. Professional School Counseling, 1(1) 43-46.
Doxsee, D.J. & Kivlighan, D.M. (1994). Hindering events in interpersonal
relations groups for counselor trainees. Journal of Counseling and Development, 72,
621-626.
Efstation, J.F., Patton, M.J., & Kardash, C.M. (1990). Measuring the working
alliance in counselor supervision. Journal of Counseling Psychology, 37, 322-329.
Ellis, M. V. & Douce, L.A. (1994). Group supervision of novice clinical
supervisors: Eight recurring issues. Journal of Counseling & Development, 72(5)
520-525.
Ellis, M.V., Dell, D.,M. & Good, G.E. (1988). Counselor trainee's perceptions
of supervisor roles: two studies testing the dimensionality of supervision. Journal of
Counseling Psychology, 35, 315-324.
Emerson, S. (1996). Creating a safe place for growth in supervision.
Contemporary Family Therapy: An International Journal, 18(3) 393-403.
Evans, N.J. & Jarvis, P.A. (1986).The Group Attitude Scale: A measure of
attraction to group. Small Group Behavior, 17, 203-216.
Friedlander, M. L. & Ward, L. G. (1984).Development and validation of the
Supervisory Styles Inventory. Journal of Counseling Psychology, 31(4) 541-557.
Fukuyama, M.A. (1994). Critical incidents in multicultural counseling
supervision: A phenomenological approach to supervision research. Counselor
Education & Supervision, 34(2) 142-151.
89
Gladding, S.T. (1991). Group work: A counseling specialty. New York:
McMillan Publishing Company.
Glidden, C. E. & Tracey, T. J. (1992). A multidimensional scaling analysis of
supervisory dimensions and their perceived relevance across trainee experience
levels. Professional Psychology: Research & Practice, 23(2) 151-157.
Grimm, L. G. & Yarnold, P.R. (1998). Reading and understanding
multivariate statistics. Washington D.C.: American Psychological Association.
Hansen, J.C., Warner, R.W., & Smith, E.J. (1980). Group counseling: Theory
and practice (2nd ed.) Chicago: Rand McNally.
Harvill, R., West, J., Jacobs, E. E., Masson, R. L. (1985). Systematic group
leader training: Evaluating the effectiveness of the approach. Journal for Specialists
in Group Work, 10(1) 2-13.
Harvill, R., Masson, R. L., Jacobs, E. E. (1983). Systematic group leader
training: A skills development approach. Journal for Specialists in Group Work, 8(4)
226-232.
Haase, R.F. & Ellis, M.V. (1987). Multivariate analysis of variance. Journal
of Counseling Psychology, 34, 404-413.
Hayes, R.L. (1990). Developmental group supervision. Journal for Specialists
in Group Work, 15(4) 225-238.
Hayes. R.L., Blackman, L.S., & Brennan, C. (2001). Group supervision. In L.
Bradley and N. Ladany (Eds) Counselor supervision: Principles. process. and
practice (3rd ed.), 183-206.
Heppner, P. P.; Roehlke, H. J. (1984). Differences among supervisees at
different levels of training: hnplications for a developmental model of supervision.
Journal of Counseling Psychology, 31(1) 76-90.
Heppner, P.P. & Kivlighan, D.M. (1994). Dimensions that characterize
supervisor interventions delivered in the context of live supervision of practicum
counselors. Journal of Counseling Psychology, 41(2), 277-235.
Hillerbrand, E. (1989). Cognitive differences between experts and novices:
hnplications for group supervision. Journal of Counseling & Development, 67(5)
293-296.
Holloway, E.L., & Johnston, R. (1985).Group supervision: widely practiced
but poorly understood. Counselor Education and Supervision, 24, 332-340.
90
Holloway, E.L., & Neufeldt, S.A. Supervision: Its contribution to treatment
efficacy. Journal of Consulting and Clinical Psychology, 63(2), 207-213.
Holmes, G.R., Stader, S.R., Swaim, K.F., Haigler, E.D., & Myers, D. (1998).
Adolescent group psychotherapy supervision in a group format: An emerging model.
Journal of Child and Adolescent Group Therapy, 8(4), 197-206.
Jennings, L. & Skovholdt, T.M. (1999). The cognitive, emotional, and
relational characteristics of master therapists. Journal of Counseling Psychology, 46
(1),3-11.
Jolliffe, I.T. (1986). Principle component analysis. New York: Springer
Verlag.
Kaiser, H.F. (1974). An index of factorial simplicity. Psychometrica, 39, 31-
36.
Kaplan, I.T. & Greenbaum, H.H. (1989). Measuring work group effectiveness:
A comparison of three instruments. Management Communication Qyarterly, 2(3),
424-448.
Kivlighan, D. M. & Quigley, S. T. (1991). Dimensions used by experienced
and novice group therapists to conceptualize group process. Journal of Counseling
Psychology. 38(4) 415-423.
Kruger, L. J., Cherniss, C., Maher, C. A., Leichtman, H. M. (1988). Group
supervision of paraprofessional counselors. Professional Psychology: Research &
Practice, 19(6) 609-616.
Ladany, N., Hill, C.E., Corbett, M.M., & Nutt, E.A. (1996). Nature, extent, and
importance of what psychotherapy trainees do not disclose to their supervisors.
Journal of Counseling Psychology, 43, 10-24.
Lanning, W. L. (1971). A study of the relation between group and individual
counseling supervision and three relationship measures. Journal of Counseling
Psychology. 18(5) 401-406.
Larson, D.L., Attkinsson, C.C., Hargreaves, W.A., & Nguyen, T.D. (1979)
Assessment of client/patient satisfaction: development of a general scale. Evaluation
and Program Planning, 2, 197-207.
Lewin, K. (1944).The dynamics of group action. Educational Leadership, 1,
195-200.
91
MacKenzie, K. R. (1981 ). Measurement of group climate. Journal of Group
Psychotherapy, 31, 287-296.
MacKenzie, K. R. The clinical application of a group climate measure .. In R.
R. Dies & K. R. MacKenzie (Eds.) Advances in group psychotherapy: Integrating
research and practice. Madison, CT: International Universities Press, 159-170.
MacKenzie, K. R., & livesley, W. J. (1986). Outcome and process measures in
brief group psychotherapy. Psychiatric Annals, 16(12) 715-720.
Marshall, R.J. (1999). Facilitating cooperation and creativity in group
supervision. Modem Psychoanalysis, 24(2), 181-186.
McAuliffe, G. J. (1992). A case presentation approach to group supervision for
community college counselors. Counselor Education & Supervision, 31(3) 163-174.
Moos, R. F. (1994). Group Environment Scale Manual. Palo Alto: Consulting
Psychologists Press.
Moos, R.F. (1974). Evaluating treatment environments. New York: John
Wiley.
Moss, E. (1995). Group supervision: Focus on countertransference.
International J oumal of Group Psychotherapy, 45( 4) 537-548.
Norman, G. R., and D. L. Streiner (1994). Biostatistics: The bare essentials.
St. Louis: Mosby.
Priest, R. Minority supervisor and majority supervisee: Another perspective of
clinical reality. Counselor Education & Supervision, 34(2) 152-158.
Prieto, L. R. (1997). Separating group supervision from group therapy.
Professional Psychology: Research & Practice, 28(4) 405.
Prieto, L. R. (1996). Group supervision: Still widely practiced but poorly
understood. Counselor Education & Supervision, 35(4) 295-307.
Rabinowitz, F.E., Heppner, P.P., & Roehlke, H.J. (1986). Descriptive study of
process outcome variables of supervision over time. Journal of Counseling
Psychology, 33, 292-300.
92
Ray, D. & Altekruse, M. (2000). Effectiveness of group supervision versus
combined group and individual supervision. Counselor Education & Supervision,
40(1) 19-30.
Riva, M. T. & Cornish, J. A. (1995). Group supervision practices at
psychology predoctoral internship programs: A national survey. Professional
Psychology: Research & Practice, 26(5) 523-525.
Sansbury, D.L. (1982). Developmental supervision from a skill perspective.
Counseling Psychologist, 10(1), 53-57.
Savickas, M.L., Marquart, C.D., & Supinski, C.R. (1986). Effective
Supervision in groups. Counselor Education and Supervision, 26(1) 17-25.
Schutz, W.C. (1958). FIRO: A three dimensional theory of interpersonal
behavior. New Y ork:Holt, Rinehart, and Winston.
Starling, P. V. & Baker, S. B. (2000) Structured peer group practicum
supervision: Supervisees' perceptions of supervision theory. Counselor Education &
Supervision, 39(3) 162-176.
Stockburger, D.W. (1997). Multivariate statistics: Concepts, Models, and
Applications. Web version: www.pschystat.smsu.edu.
Stoltenberg, C.D., McNeill, B.W., & Crethar, H.C. (1994). Changes in
supervision as counselors and therapists gain experience: a review. Professional
Psychology: Research and Practice, 25(4), 416-449.
Stoltenberg, C.D., Pierce, R.A., & McNeill, B.W. (1987). Effects of experience
on counselor trainee's needs. The Clinical Supervisor, 5(1), 23-32.
Stoltenberg, C.D. & Delworth, U. Supervising counselors and threapists. San
Francisco: Jossey-Bass.
Swanson, J. L. & O'Saben, C. L. (1993). Differences in supervisory needs and
expectations by trainee experience, cognitive style, and program membership.
Journal of Counseling & Development, 71(4) 457-464.
Tabachnick, B. G. & Fidell, L.S. (1996). Using multivariate statistics. New
York, NY: Harper Collins.
Tebb, Susan; Manning, Donald W.; Klaumann, Teresa K. A renaissance of
group supervision in practicum. Clinical Supervisor. 1996 Vol14(2) 39-51.
93
Tuckman, B.W. & Jensen, M.C. (1977). Stages of small group development
revisited. Group and Organizational Studies, 2, 419-427.
Walter, C.A. & Young, T. M. (1999). Combining individual and group
supervision in educating for the social work profession. Clinical Supervisor, 18(2)
73-89.
Wampold, B.E. & Holloway, E.L. (1997) Causal model of supervision. In C.E.
Watkins' (Ed.) Handbook of psychotherapy supervision. New York: John Wiley &
Sons.
Watkins, C.E. (1997). Defming psychotherapy supervision and understanding
supervisor functioning. In C.E. Watkins' (Ed.) Handbook of psychotherapy
supervision. New York: John Wiley & Sons.
Westwood, M. J. (1989). Group supervision for counselors-in-training.
Canadian Journal of Counseling, 23(4) 348-353.
Wheeler, J.L. & Kivlighan, D.M. (1995).Things unsaid in group counseling:
An empirical taxonomy. Journal of Counseling and Development, 73, 586-590.
Wilbur, M.P., Roberts-Wilbur, J., Hart, G. M., Morris, J. R. (1994). Structured
Group Supervision (SGS): A pilot study. Counselor Education & Supervision, 33(4)
262-279.
Wilbur, M.P., Roberts-Wilbur, J., Morris, J. R., Betz, R. L. (1991) Structured
group supervision: Theory into practice. Journal for Specialists in Group Work,
16(2) 91-100.
Worthington, E. L. (1981). Empirical investigation of supervision of
counselors as they gain experience. Journal of Counseling Psychology, 31(1) 63-75.
Yalom, I. (1995). The theory and practice of group psychotherapy. (4th
Edition). New York: Basic Books.
94
Appendix A
Group Supervision Scale
The Development and Validation of the Group Supervision Scale
You are asked to participate in a research study conducted by Ferdinand
Arcinue, M.A. and Rodney Goodyear, Ph.D., from the Division of
Counseling Psychology at the University of Southern California. The
results will contribute to the completion of Ferdinand Arcinue's Ph.D.
dissertation. You were selected as a possible participant in this study
because of your involvement in group supervision. We are asking you to
take part in a research study to validate an instrument that evaluates
the performance of a group supervisor. Your participation in this study
will help provide the field with a reliable and valid instrument that
measures several aspects that are considered to be the responsibility of a
supervisor leading a group of trainees. Since there are no validated
instruments of this nature, your participation is important given the
widespread use of group supervision as a training medium. Completion
and return of the questionnaire or response to the interview questions
will constitute consent to participate in this research project.
You will be asked to spend approximately 10-15 minutes to complete this
survey including demographic items, questions about your group
supervisor, and your group supervision. This study has been reviewed by
the Institutional Review Board at USC (#02-01-008). There are no
potential negative effects, risks, and discomforts from participating in this
study. You are expected to gain no benefits or financial payments from
participating in this study. This is a web-based survey. No information
will be obtained that may reveal your identity. Your participation will be
confidential and no one will be given information about your participation
or your responses. This survey does not ask for your name. Please do not
indicate your name anywhere on it. Any information that is obtained in
connection with this study and that can be identified with you will remain
confidential and will be disclosed only with your permission or as required
by law. When the results of the research are published or discussed in
conferences, no information will be included that would reveal your
identity. Your responses will be stored on a secure server, downloaded
directly by Ferdinand Arcinue, M.A., and entered into a database seen only
by those involved with this research project. Individual responses to this
95
Appendix A continued
survey questionnaire will be destroyed within 3 months following the
analyses of the data. You can choose whether to be in this study or not. If
you volunteer to be in this study, you may withdraw at any time without
consequences of any kind. You may also refuse to answer any questions
you don't want to answer and still remain in the study. If you have any
questions or concerns about the research, please contact Rodney Goodyear,
Ph.D. at goodyea@usc.edu or (213) 740-3267. Questions or concerns may
also be mailed to Dr. Goodyear at the Division of Counseling Psychology,
WPH 503, University of Southern California, Los Angeles, CA 90089-
0031.
You may withdraw your consent at any time and discontinue participation
without penalty. You are not waiving any legal claims, rights or remedies
because of your participation in this research study. If you have questions
regarding your rights as a research subject, contact the University Park
IRB, Office of the Vice Provost for Research, Bovard Administration
Building, Room 300, Los Angeles, CA 90089-4019, (213) 740-6709 or
upirb @usc.edu.
Thank you for taking the time to participate in this study.
Please fill in, check, or circle the appropriate answer to each question.
1. Your Age:'---------:-------
2. Gender:
C Male
C Female
96
Appendix A continued
3. Ethnic Background:
G Asian or Pacific Islander
C Black, African-American
r;: Caucasian (non-Hispanic)
r;: Hispanic!Latino
c
Native American
C other (please enter) L __ _
4. Your current degree objective:
C MA
C MSW
C PhD
r::
PsyD
r;:: other (please enter) ._I ---
5. Profession or Specialty:
C Counseling Psychology
C Clinical Psychology
C Counseling (e.g. MFf)
C School Counseling
c
Social Work
C other (please enter)!._---
Optional*. What is the name of your university (please do not abbreviate)?
(this item will be used to assess participation in the study and will not be
connected to individual data)
L __ _
97
Appendix A continued
6. Year in current program:
c
1st (started in 2001)
c
2nd (started in 2000)
c
3rd (started in 1999)
c
4th (started in 1998)
c
5th (started in 199[
c
other (enter year)
7. Do you already have a master's degree?
C Yes
r::
No
8. Approximately how many semesters of supervised practicum or field
placement have you completed?
Number of
semesters
I ___ _
9. Please think of your current group supervision experience in which you
participate as a supervisee.
How many weeks has this group C
met?
How many supervisees are in your L ____ __
group?
10. What was the context in which the group occurred?
C university based practicum course
C counseling center
C community agency
C hospitaVinpatient setting
C other (please enter) r-[_- __ ...... __ '-_ ------
Please answer the following items regarding your current group
supervision experience. If you are in more than one supervision group,
please choose one to answer all of these items.
98
Appendix A continued
11. To what degree did your group supervision emphasize these domains?
Please assign a rating on a scale from 1 to 6, where 1 represents "Not at
All" and 6 represents "Always/Definitely".
1 2 3 4 5 6
..
-
Interpersonal Process (self-
c c c c c c
awareness/personal growth)
-- - ~ -·
-- --- ·-- --
Case Presentation/Discussion (presentation c c r .'· c c c
of clinical material)
Didactic Information (teaching of specific
0 c c c c . c
material/interventions)
-
12. Group Climate Measure (MacKenzie, 1981).
Please answer the following questions regarding the environment of your
supervision group. Assign a rating on a scale :from 1 to 6, where 1
represents "Not at All" and 6 represents "Always/Definitely".
- - - -
s16 1 2 3 4
The members liked and cared about each c c c 0 c C·
other.
-
The members tried to understand why they
c c c c 0 c
do the things they do, tried to reason it out
~
· The members felt what was happening was
c c c c c c
important and there was a sense of
iparticiP!tion.
- --
..
The members challenged and confronted 0 c C· c c c
each other in their efforts to sort things out
The members revealed sensitive personal
c c c 0 c c
information or feelings.
-
99
Appendix A continued
13. Supervision Satisfaction Questionnaire (Ladany, Hill, & Nutt, 1996)
Please answer the following regarding your group supervisor and group
supervision experience. Assign a rating on a scale from 1 to 6, where 1
represents "No, definitely not" and 6 represents "Yes, definitely".
- -
1 2 3 4 5 6
Did you get the kind of supervision you 0 0 c 0 c c
wanted?
-
If a friend were in need of supervision, c c c 0 0 c
would you recommend this supervisor?
Has the supervision you received helped
c c c
(i
c c
you to deal more effectively in your role as
a counselor or ~rapist?
- - -
If you were to seek supervision again, c 0 0 0 c 0
would you come back to this supervisor?
13B. Supervision Satisfaction Questionnaire continued
Please assign a rating on a scale from 1 to 6, where 1 represents "Low" and
6 rep t "H" h" resen s IgJ
1 2 3 4 5 6
How would you rate the quality of the c 0 c c 0 0
supervision you have received?
-
To what extent has this supervision fit your 0 c 0 c c c
needs?
. -·
How satisfied are you with the amount of
c .· c c 0 c c
supervision you have received?
In an overall, general sense, how satisfied
0 c c 0 c c
are you with the supervision you have
received?
-- - --
100
Appendix A continued
14. Supervisory Working Alliance Inventory (Efstation, Patton, & Kardas~
1990)
Please assign a rating on a scale from 1 to 6, where 1 represents "Almost
Never" and 6 represents "Almost Always".
- - -
1 12 3
- ··~ -- - - -
In group supervision. my supervisor places
c c 0
a high priority on our understanding the
client's tive.
My group supervisor encourages me to take
c c 0
time to understand what the client is saying
~d ~~in_g_: -
. - . - . ·- - -
..
My group supervisor's style is to carefully
c c 0
and systematically consider the material I
bring to supervisio~.
When correcting my errors with a client.
c c c
my group supervisor offers alternative ways
of interyeniJ!g with that client.
My group supervisor helps me work within 0 c c
a specific treatment plan with my clients.
- --
My group supervisor helps me stay on track 0 0 0
during our meetings.
I work with my group supervisor on c c c
specific goals in the supervisory session.
Please answer the following items regarding your group supervisor.
15A. What is your group supervisor's gender:
0
Male
° Female
-·
4 5 6
c c r
c c c
-
c c c
c 0 c
c . c 0
0 c 0
c c c
101
Appendix A continued
15B. To the best of your knowledge, what is your supervisor's ethnicity?
0
Asian/Pacific Islander
0
Black/ African-American
° Caucasian (non-Hispanic)
0
Hispanic/Latino
0
Native American
0
Other (please enter) t_ __ .. ______________ j
15C. To the best of your knowledge, approximately how old is your group
supervisor?
approximate age in I ..... __ ........ __
years
16. Supervisor's Appearance
Please assign a rating on a scale from 1 to 6, where 1 represents "Not at
All" and 6 represents "Very".
1 2 3 4 5 6
-
How physically attractive is your
c c c c c c
supervisor?
How professional in appearance is your
c c c c c c
supervisor?
102
Appendix A continued
17. Group Supervision Scale
Please answer the following items regarding your group supervisor. Assign
a rating on a scale from 1 to 6, where 1 represents "Strongly Disagree" and
6 represents "Strongly Agree".
··- - -·- -·- --- --~- -· --~- - ---- - -- --- - --
1 ~ 2 3 4 5 6
- -
rJ
The supervisor provides useful feedback 0 c c c c
regarding my skills and interventions.
!
- - - - -
- .
The supervisor provided helpful
c c 0 c c
(',
suggestions and information related to
.. ·
treatment interventions.
.. -
- --
..
The supervisor facilitates constructive
c c c c c r
exploration of ideas and techniques for
.. •
I
working with clients.
The supervisor provides helpful
c c c c c 0
information regarding case
I
con~tualization and di~os!s.
- - -
:
The supervisor helps me comprehend and c c 0 c c ,C
formulate clients' central issues.
·-
...
--- ·-- -
The supervisor helps me to understand the c c c c c c
thoughts, feelings, and behaviors of clients. I
- -
_,
-
The supervisor encourages trainee self- c c c c c c
exploration appropriately.
--' J
-
The supervisor enables me to express
c c
()
c c c I
opinions, questions, and concerns about my
c~unseling.
I
_j
. .. ... ...
. The supervisor created a safe environment c c 0 c c c
for group supervision.
! l
... .....
-
..
- -
The supervisor is attentive to group c c c
I
c c ·C
dynamics.
:
-- -
...
--
....
- - --J --
The supervisor effectively sets limits, and
!
! establishes norms and boundaries for the
c c c c
,c
c
.group.
I
The supervisor provides helpful leadership c c c c c c I
for the group. I
103
Appendix A Continued
The supervisor encourages supervisees to
c c r. 0 G G
provide each other feedback.
The supervisor redirects the discussion
c
()
c c c c
when appropriate.
The supervisor manages time well between
c 0 n c c
·~ n
all the group members.
-
The supervisor provides enough structure in c 0 c 0 c c
the group supervision.
Thank you very much for your participation!
104
AppendixB
Histograms of GSS, GCQ, SWAI, and SSQ.
Group Supervision Scale (GSS)
3Qr-------------------------------.
20
10
0
20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0
25.0 35.0 45.0 55.0 65.0 75.0 85.0 95.0
GSS.F
Group Climate Questionnaire (GCQ)
50
'
30
20
10
0
10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5 30.0
GCM.F
Std. Dev = 1827
Mean=70.3
N=226.00
Std. Dev = 4.53
Mean=21 .4
N=226.00
105
Appendix B continued
30
20
10
0
Supervisory Working Alliance Inventory (SWAI)
"client focus" subscale
7.5 12.5 17.5 22.5 ZT.S 32.5 37.5 42.5
10.0 15.0 20.0 25.0 30.0 35.0 40.0
SWAI.F
Supervision Satisfaction Questionnaire
Std. Dev = 7.73
Mean=28.7
N=226.00
40~----------------------------~
30
20
10
0
7.5 12.5 17.5 22.5 ZT.S 32.5 37.5 42.5 47.5
10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0
SSQ.F
Std. Dev = 10.46
Mean=35.1
N=226.00
106

Asset Metadata

Creator Arcinue, Ferdinand (author)

Core Title The development and validation of the group supervision scale

Contributor Digitized by Interlibrary Loan Department (provenance)

Degree Doctor of Philosophy

Degree Program Education (Counseling Psychology)

Defense Date 12/01/2002

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag OAI-PMH Harvest

Format application/pdf (imt)

Language English

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-490018

Unique identifier UC11286576

Identifier etd-Arcinue-583791.pdf (filename),usctheses-c3-490018 (legacy record id)

Legacy Identifier etd-Arcinue-583791.pdf

Dmrecord 490018

Document Type Dissertation

Format application/pdf (imt)

Rights Arcinue, Ferdinand

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA

Linked assets

University of Southern California Dissertations and Theses