Content uploaded by James Herbert
Author content
All content in this area was uploaded by James Herbert on Jun 17, 2014
Content may be subject to copyright.
Evaluation of a Training Program to Enhance
Clinical Supervision of State Vocational
Rehabilitation Supervisors
James T. Herbert
Soo-yong Byun
Penn State University
Jared C. Schultz
Utah State University
Meritza Tamez
Heather A. Atkinson
Penn State University
Herbertetal.
Abstract. Clinical supervision is intended to improve counseling skills and case management
decisions made by the rehabilitation counselor that impact the client-counselor relationship
and ultimately vocational rehabilitation outcome. We evaluated the effectiveness of a training
program for state vocational rehabilitation supervisors to improve their clinical supervision.
The training program involved a 3-day on-site foundational training component followed by a
six-month, bi-weekly consultation supervision session (approximately 75 minutes each) to further
develop clinical supervision competence. A repeated measures design using four outcome
measures (clinical supervision knowledge, behavior, self-efficacy, and working alliance) across
baseline, initial post-test (6 weeks after the on-site training), and second post-test (6 months after
baseline) were used to determine program effectiveness. Parallel measures for each of the four
outcome variables were administered on-line to supervisors who completed the initial training as
well as counselors who completed outcome measures at baseline (n=21 supervisors; n=73
counselors), post-test 1 (n=17 supervisors; n=37 counselors) and post-test 2 (n=12 supervisors;
n=24 counselors) administrations. Results indicated that supervisors reported a continual
increase in clinical supervision knowledge across the three observations periods. No changes
were found with regard to clinical supervision behavior, self-efficacy and working alliance,
however. Further, no statistically significant change occurred in any outcome measure reported
by counselors. Results are discussed in terms of the research questions as well as strategies to
advance clinical supervision research.
Keywords: clinical supervision, vocational rehabilitation
Within the state vocational rehabilitation (SVR) system
there are two aspects of counselor supervision both of
which are designed to improve service delivery and
successful outcomes — administrative and clinical
Journal of Rehabilitation Administration, 38(1), pp. 19–34, 2014
supervision (Herbert, 2003). Administrative supervi-
sion concerns the documentation of vocational reha-
bilitation efficiency and effectiveness of services
provided by rehabilitation counselors and related
personnel matters in the discharge of these profes-
sional responsibilities. Within the SVR system, this
form of supervision focuses on counselor perfor-
mance indicators such as length of time clients typi-
cally require between initial referral for services
(Status 02) and an approved individualized written
rehabilitation plan is completed (Status 12), number
of successful competitive closures achieved by the
counselor (Status 26) and amount of case dollar ex-
penditures within a particular time frame. These indi-
cators pertain to service efficiency and effectiveness
which have more of an agency focus related to coun-
selor performance (Herbert, 1995). Typically, they
are conducted after service delivery and evaluated
through written case documentation and related per-
formance indicators that are typically reviewed by
some predetermined time schedule (e.g., every three
months). This type of supervision does not, however,
fully address an important component that impacts
service delivery, namely the client-counselor rela-
tionship.
Clinical supervision, in contrast, addresses a differ-
ent but complementary aspect. This aspect involves
an evaluative yet supportive relationship between
counselor and supervisor, where the focus centers on
counselor-client relationship dynamics to promote
counselor skill development and effective case man-
agement decisions (Herbert & Caldwell, in press).
Examples of clinical supervision may involve an ex-
amination of a counselor’s attitudes toward specific
clientele and its impact on the client-counselor rela-
tionship, uncertainty about addressing employer ob-
jections regarding hiring persons with disabilities or
listening more empathically when client concerns are
not overtly expressed during an intake interview. As
these examples demonstrate, the clinical supervisor
within SVR practice can examine a variety of aspects
that impact job performance but, most often, those
that involve the client-counselor relationship and
how to promote an effective working alliance that re-
sults in successful vocational rehabilitation outcomes
(Herbert, 2009). The importance of examining cli-
ent-counselor relationship issues as part of supervi-
sion is particularly relevant given the consistent body
of literature supporting how this relationship contrib-
utes to successful vocational rehabilitation outcome
(e.g., Bozarth et al., 1975; Capella-McDonnall, 2005;
Donnell, Lustig, & Strauser, 2004; Garske & Soriano,
1997; Kissinger, 2008; Lustig, Strauser, Rice, & Rucker,
2002; Strauser, Lustig & Donnell, 2004; Thomas &
Whitney-Thomas, 1996).
Clinical Supervision Relevance to
Counseling Practice
An examination of the published literature within the
past 30 years indicates relatively few studies have
specifically investigated supervisory practices within
the SVR system. For example, in the first and only
nationwide study of counselors employed in the pub-
lic rehabilitation sector, English, Oberle, and Bryne
(1979) found relevant to this discussion that SVR su-
pervisors (a) devoted most of their efforts towards
administrative tasks and case consultation tasks, (b)
primarily used case review and field observation
methods when supervising counselors, and (c) did a
“poor job in many functional areas of counselor con-
sultation and counselor evaluation” (p. 11). A final
conclusion from the study was that although SVR
counselors were critical of all aspects of supervision
they believed that with additional training these skills
could improve.
More recently, Schultz, Ososkie, Fried, Nelson, and
Bardos (2002) examined supervision practices as
well as the supervisory relationship within the SVR
program from two states. Major findings pertinent to
this review found the following: (a) Most supervision
provided to VR counselors occurred on an as needed
basis and, as a result, “eliminat[ed] the opportunity
for engaging in proactive activities focused on coun-
selor development” (p. 219); (b) Fewer than 1 in 3
counselors reported having a regularly scheduled
time to meet with the supervisor; and (c) There was a
“severe lack of understanding” regarding the pur-
pose, practice, and potential benefits of clinical su-
pervision.
In assessing supervisory needs and clinical supervi-
sion practices within the SVR system, Herbert
(2004a) conducted a study of 148 Pennsylvania VR
supervisors and counselors and found that the major-
ity of participants either had an uninformed or gen-
eral misunderstanding of the nature of clinical super-
vision. Of those who understood how it could be used
in SVR practice, they tended to be new counselors
(those with two years experience or less) who gradu-
ated from a master’s degree program in rehabilitation
counseling. Interview data also revealed that clinical
supervision is an activity that rarely occurred with
experienced counselors but occurred somewhat more
frequently among newer counselors (in about 20% of
Journal of Rehabilitation Administration, 38(1), 2014
Page 20 Herbert et al.
instances). When asked about major obstacles in pro-
viding or receiving effective clinical supervision,
three barriers were identified: (a) insufficient time to
perform clinical supervision, (b) lack of training in
clinical supervision, and (c) counselor-supervisor
dynamics that result in uncomfortable professional
relationships. Perhaps not surprisingly, when asked
to provide recommendations to improve clinical su-
pervision practice, the most frequent suggestion was
to provide training to supervisors on how to imple-
ment clinical supervision.
In a subsequent companion study, Herbert and Trusty
(2006) found that among supervisors (n = 39) and
counselors (n = 104), the average time devoted to
both administrative and clinical supervision was
about 20 minutes each month. Only 20% of partici-
pants either provided or received supervision each
week whereas nearly 80% had supervision provided
on a monthly basis or longer. When supervision oc-
curred, in 93% of instances, the rehabilitation coun-
selor initiated it rather than any pre-established regu-
larly scheduled supervision initiated by the supervi-
sor. In addition, 90% of supervisors had not received
any formal training in clinical supervision. In terms
of supervision formats, although 35% of supervisors
indicated that they used both individual and group
supervision formats, no counselor ever reported re-
ceiving group supervision. This apparent discrep-
ancy seemed attributable to supervisors believing
that unit meetings where general agency and related
information was shared were synonymous with clini-
cal group supervision.
Despite ignoring the need for clinical supervision
training documented over 30 years ago by English et
al. (1979), and what Olshansky (1973) characterized
as a “scandal in vocational rehabilitation,” clinical
supervision is an important activity that both SVR
counselors recognize and need yet, at the same time,
perceive limited competence (Herbert, 2004a; Her-
bert & Trusty, 2006; Schultz et al., 2002). The pur-
pose of this study was to develop an intervention de-
signed to increase clinical supervision knowledge,
enhance the belief that one can effectively provide
clinical supervision (efficacy), increase frequency of
effective clinical supervision behaviors and
strengthen the working alliance between counselors
and supervisors within the context of clinical super-
vision. The associated research questions within this
purpose were as follows:
Would a statistically significant increase occur
across these four outcome measures as observed at
baseline, initial post-test (Time 1 – six weeks later)
and second post-test (Time 2 – six months after Base-
line) among supervisors who participated in the
training program (i.e., Time 2 ³Time 1³Baseline)?
Similarly, would counselors assigned to these super-
visors who participated in the training program re-
port a statistically significant change among the four
parallel outcome measures across the three observa-
tion periods (i.e., for each measure Time 2 ³Time 1 ³
Baseline)?
Would an increase in frequency, duration and quality
of individual and group supervision occur as reported
by counselors and supervisors across the three obser-
vation periods (i.e., Time 2 ³Time 1³Baseline)?
Would the experimental scales measuring clinical
supervision behavior, efficacy and knowledge dem-
onstrate sufficient reliability (i.e., internal consis-
tency estimates of .90 or higher)?
Methodology
In developing the training program to enhance clinical
supervision competence, content domains were ini-
tially identified from studies conducted in the field that
identified effective and less effective behavior and
perceived training need areas within SVR practice
(e.g., Herbert, 2004a; Theilsen & Leahy, 2001) as well
as that identified in other counselor supervision litera-
ture (Bernard & Goodyear, 2004; Worthington &
Roehlke, 1979). An initial list of potential topics was
then evaluated by a group of eight experts from six
states who served as part of a research and training ad-
visory board. Each of these members had considerable
experience (25 years, on average) who worked in the
state vocational rehabilitation system including ad-
ministrators, training directors and supervisors. A pre-
liminary list of 95 topics was developed using several
resources including research studies (e.g., Herbert,
2004b; Schultz et al., 2002), reference material (e.g.,
Bernard & Goodyear, 2004; Borders, 1989; Herbert,
2004c; Schultz, 2008), existing (Thielsen & Leahy,
2001) and experimental clinical supervision scales
that contained perceived content topics relevant to
state VR supervision practice (e.g., Stebnicki, Allen,
& Janikowski, 1997). The list of topics followed these
content domains: (a) differentiating administrative
from clinical supervision, (b) obstacles to supervision
practice and best practices, (c) counselor and supervi-
sor developmental models, (d) supervisory roles (con-
sultant, counselor, and teacher) used in supervision
and methods to promote the supervisory working alli-
ance, (e) effective practices when providing individual
Journal of Rehabilitation Administration, 38(1), 2014
Clinical Supervision Training Page 21
and group supervision, (f) evaluation techniques and
process, and (g) multicultural aspects of clinical super-
vision. From this list, experts rated each topic in terms
of its relevance to include as part of the clinical super-
vision training program being developed. After view-
ing group mean responses for each item, the Advisory
Board then went through a second round of previously
assessed topics and decided which topics would be re-
tained for the training program. This process formed
the content as part of the initial 3-day foundational
training given to SVR supervisors with standardized
content, training activities and supplemental materi-
als. Following this initial training, a six-month fol-
low-along consultation program that met bi-weekly
via teleconferencing (telephone) for approximately 75
minutes with supervisors who participated in the foun-
dational program was initiated. This follow-up train-
ing built on foundational skills developed in the initial
program. The first investigator facilitated both foun-
dational training and follow-along training. Because
of group size (n = 22), the follow-along training was
divided into three groups with about seven members
in each group and, in this way, each supervisor would
have sufficient opportunity to address clinical supervi-
sion concerns he/she wanted to discuss with their
peers. Group assignment was based solely on avail-
ability of times that matched the availability of other
group members to meet. When desired by the group,
supplemental readings related to clinical supervision
and all of the training materials were placed on a
website specifically created for the training program.
Supervisors then had opportunity to review earlier
material and, if interested, review supplemental mate-
rials at any time they wished to do so. The primary in-
vestigator monitored session notes of follow-along
session content (i.e., general themes) as well as partici-
pant attendance.
Outcome Measures
Assessing training effectiveness was evaluated using
four outcome measures pertaining to clinical supervi-
sion behavior, knowledge, self-efficacy and working
alliance between counselors and supervisors. Other
behavioral indicators regarding frequency and length
of individual and group supervision as well as overall
satisfaction with supervision provided or received was
also ascertained across three observation periods. This
multi-method and multi-source strategy has often
been touted as characteristic of strong research design
but one that is not used very often in social sciences re-
search (Tabachnick & Fidell, 2001) and those pertain-
ing to vocational rehabilitation research studies
(Saunders, Leahy, McGlynn, & Estrada-Hernandez,
2006).
Clinical supervision behavior. Items from this ex-
perimental scale were largely based on qualitative in-
terviews of SVR counselors and supervisors con-
ducted by Herbert (2004a) who identified effective
clinical supervision behavior (e.g., maintain regu-
larly scheduled meetings, accompany counselors in
the field to observe client-counselor interactions, fol-
low through on professional commitments, use
role-play during supervision, incorporate case pre-
sentations as part of group supervision, provide re-
sources that may help counselors improve as a coun-
selor [e.g., journal articles, information found on
Internet]). The Likert-type scale was comprised of 29
items with each item asking the level of agreement
pertaining to a behavioral indicator noted in actual
practice ranging from “1 = Strongly Agree” to “6 =
Strongly Disagree.”
Clinical supervision knowledge. An adapted scale
based on the Rehabilitation Counseling Supervision
Inventory (Thielsen & Leahy, 2001) was developed
using four of the six knowledge subscales—Ethical
and Legal Issues, Intervention Techniques and Meth-
ods, Assessment and Evaluation, and Supervisory
Relationship—as the initial item pool (n = 63). These
items were then evaluated by advisory board mem-
bers in terms of their relevance to training program
content and modified to better fit question presenta-
tion format. For example, an original item “Rapport
building in supervision” was rewritten as “Build rap-
port with a counselor during supervision.” In addi-
tion, several new items were created to reflect infor-
mation advisory board members deemed relevant to
the training program that they did not perceive con-
tained in the initial scale by Thielsen and Leahy. This
process resulted in a 33-item scale to assess clinical
supervision knowledge. Using the same response
format as the original scale, each item used a 5-point
Likert-type scale ranging from 0 = “No knowledge or
understanding” to 4 = “Very high degree of knowl-
edge or understanding.” Participants could also se-
lect “U” to indicate “Unsure or can not evaluate.”
Clinical supervision self-efficacy. Self-efficacy is a
belief in one’s capabilities to produce a given out-
come (Bandura, 1997). As a situational construct,
self-efficacy is differentiated by a set of self-beliefs
linked to specific behaviors and situations (Bandura,
2006). As a result, a measure specifically related to
clinical supervision self-efficacy had to be devel-
oped. Since no instruments existed to assess this con-
Journal of Rehabilitation Administration, 38(1), 2014
Page 22 Herbert et al.
struct, an experimental measure with 15 items was
developed. In offering specific suggestions on how to
construct self-efficacy scales Bandura (2006) pro-
vided the following guidelines:
1. Items should be phrased in “can do” rather than
“will do” (as the former is a judgment in capa-
bility).
2. Items should address how well a person can per-
form in the present situation.
3. Efficacy scales should be multifaceted so that it
provides insights into the dynamics of the be-
havior.
4. Efficacy scales must be tailored to activity do-
mains and linked to judgment immediately be-
fore each performance.
5. Efficacy statements should be free of social de-
sirability or other responding bias.
Incorporating these guidelines as a way to develop a
research scale on supervisor self-efficacy, three ma-
jor roles and associated behaviors that are often de-
scribed in clinical supervision—consultant, coun-
selor, and teacher roles (Bernard & Goodyear, 2004)
served as the initial framework. Using an earlier
model articulated by Stenack and Dye (1982) based
on these three major roles, an experimental measure
was developed. This measure contained five state-
ments associated with each of the three roles where
respondents were asked to rate their confidence level
(from 0 [can not do at all] to 100 [highly certain can
do]). Items included these: “Teach, demonstrate, or
model counseling intervention techniques”, “Inter-
pret significant events in the counseling session” and
“Process professional concerns that I [counselor]
might get defensive about.”
Supervision working alliance. One of the most fre-
quently used scales in clinical supervision is the Su-
pervisory Working Alliance Inventory (SWAI; Ber-
nard & Goodyear, 2004). Developed by Efstation,
Patton and Kardash (1990), the SWAI measures the
supervisory relationship as to how supervisors act to
influence supervisee knowledge and skill and how
supervisees reciprocate in their display of acquisition
of knowledge and skill. There are two forms of the
SWAI, one that measures satisfaction of the supervi-
sory relationship from the perspective of the trainee
or the person being supervised, which is the Supervi-
sory Working Alliance – Trainee (SWAI-T) form (19
items). Examples of items include these: “My super-
visor helps me to talk freely in our sessions”, “My su-
pervisor treats me like a colleague in our supervisory
sessions” and “My supervisor stays in tune with me
during supervision.” A parallel assessment that mea-
sures satisfaction from the perspective of the supervi-
sor, Supervisory Working Alliance – Supervisor
(SWAI-S) containing 23 items, was also used. Exam-
ples of test items include these: “I help my counselors
stay on track during meetings”, “My counselors ap-
pear to be comfortable working with me” and “I
teach my counselors through direct suggestion.” For
both versions, test items are worded on a 7-point
Likert-type measure where responses can range from
“almost never” (1) to “almost always” (7). Scores for
the trainee form could range from values between 19
and 133 whereas scores for the supervisor form that
contains four additional items could range from 23 to
161. In each case, higher scores indicate a stronger,
more effective supervisor relationship as perceived
by either trainee (counselor) or supervisor.
Similar to the rationale of using parallel forms of the
supervisory working alliance, there was a need to de-
velop, a priori, equivalent experimental forms from
both counselor and supervisor perspectives for each
of the remaining three outcome measures as it per-
tained to clinical supervision behavior, knowledge
and self-efficacy. For example, an item on the clini-
cal supervision behavior scale supervisor form read,
“Even though an appointment was not scheduled, I
still made sufficient time to meet with counselors to
discuss client issues and related concerns” whereas
the counselor form read, “My supervisor made suffi-
cient time available to meet with me about a client or
related concern when there was no prior scheduled
appointment.” A similar format was adopted for the
clinical supervisor knowledge scale. For example, an
item on the supervisor form read, “I had a tendency to
‘take over’ during supervision” compared with the
counselor form that read, “[My supervisor] has a ten-
dency to ‘take over’ during supervision.” As it per-
tained to self-efficacy, respondents used a scroll bar
that allowed them to rate their confidence level on a
scale of 0 (“Not at all confident”) to 100 (“Highly
confident”) when performing or observing a specific
task. For example, for each item, a stem phrase on the
supervisor form would begin with, “Right now, I feel
thatIcaneffectivelycan....”followedbyaspecific
efficacy item such as, “. . . address and facilitate wor-
ries the counselor might have about the counseling
session.” The parallel counselor item read, “Right
now, I feel that my supervisor can effectively address
and facilitate worries I might have about the counsel-
ing session.”
Journal of Rehabilitation Administration, 38(1), 2014
Clinical Supervision Training Page 23
Psychometrics of Dependent Measures. As noted
in Table 1, internal consistency estimates taken at
baseline for both counselor and supervisors samples
were quite high, indicating that the within-scale
items associated with clinical supervision behavior,
efficacy, knowledge and working alliance were
highly correlated with one another. Given that each
measure was intended to examine some aspect of su-
pervision pertaining to beliefs, knowledge, self-effi-
cacy and working alliance between counselor and su-
pervisor, as expected, inter-correlations among the
four outcome measures were also moderately to very
high (see Table 1).
Procedure
After an initial consultation and approval from the
administrative team with the state agency, an on-site
presentation explaining the purpose, experimental
procedures and expected benefits from participating
in the study was made to supervisors from a mid-At-
lantic state. A follow-up electronic letter reviewing
the presentation content and inviting supervisors to
participate was sent afterwards to all VR supervisors
(N = 24) and, of these, 21 or 88% agreed to partici-
pate in the clinical supervision training program. A
similar letter was also sent to VR counselors inviting
them to participate in the study. Of the 79 counselors
eligible to participate, 73 or 92% participated at the
time when baseline data were collected.
Data were collected on three occasions—approxi-
mately one month prior to the three-day on-site train-
ing (baseline), about six weeks after the training
(post-test 1) and five months later (post-test 2). In or-
der to maintain confidentiality, each participant was
assigned an individualized link with an access code
only known to her or him and the lead investigator.
When taking the on-line assessment, participants
read an informed consent narrative and, if they pro-
ceeded, consent was inferred. With the exception of
demographic information that was collected at base-
line, subsequent administrations of the outcome in-
struments remained the same. In addition to items
pertaining to the four outcome assessments, addi-
tional questions were asked regarding the frequency
and length of individual and group supervision
within the past 30 work days. Participants were also
asked to comment on the level of satisfaction they
perceived when either giving (supervisors) or receiv-
ing (counselors) clinical supervision. Each partici-
pant was assigned an individual participant code that
allowed tracking of each person’s responses through-
out the study.
When solicitations to collect on-line data were made,
participants received two follow-up email notices
asking for their continued participation. As an incen-
tive for participating, persons received continuing
education units (CEUs) approved by the Commis-
sion on Rehabilitation Counselor Certification when
they completed the on-line assessment. Supervisors
received additional CEUs for participating in the
three-day on-site instruction that was provided.
Journal of Rehabilitation Administration, 38(1), 2014
Page 24 Herbert et al.
Table 1
Inter-correlations and Reliability Estimates of Outcome Measures*
Scale Internal
Items Behavior Efficacy Knowledge Consistency
Supervisor (n=21)
Behavior 29 — — — .94
Efficacy 15 .75 — — .96
Knowledge 33 .73 .86 — .96
Working Alliance .23 .61 .79 .73 .93
Counselor (n=73)
Behavior 29 — — — .96
Efficacy 15 .79 — — .98
Knowledge 33 .83 .87 — .98
Working Alliance 19 .78 .88 .86 .97
Note. *All correlations were significant at .01 level (2-tailed)
Journal of Rehabilitation Administration, 38(1), 2014
Clinical Supervision Training Page 25
Table 2
Demographics
Variable Counselor Supervisor
Mean SD Mean SD
Age 43.54 12.90 50.33 10.08
Experience as VR Counselor 9.74 9.09 13.68 9.58
Experience as VR Supervisor — — 8.33 6.89
Client Caseload 127.10 40.72 32.43 41.15
Disability Status n % n %
None 53 72.6 16 76.2
Hearing 1 1.4 2 9.5
Vision 1 1.4 1 4.8
Orthopedic 9 12.3 1 4.8
Other 3 4.1 0 0.0
No answer 6 8.2 1 4.8
Ethnicity
Black/African American 21 28.8 6 28.6
Latino(a)/Hispanic 2 2.7
Native Hawaiian/Pacific 1 1.4
White/Caucasian 41 56.2 15 71.4
Other 2 2.7
No Answer 6 8.2
Gender
Female 47 64.4 15 71.4
Male 20 27.4 6 28.6
No answer 6 8.2
Educational Level
Bachelor’s 19 26.0 4 19.0
Master’s 52 71.2 17 81.0
Doctorate 2 2.7 0 0.0
Degree Area
Rehabilitation Counseling 29 39.7 10 47.6
Other Counseling Area 11 15.1 2 9.5
Psychology 11 15.1 3 14.3
Social Work 1 1.4 1 4.8
Other 21 28.8 5 23.8
Certification/Licensure
CRC Only 11 15.1 6 28.6
LPC Only 1 1.4 0 0.0
More than one 4 5.5 0 0.0
None 53 72.6 15 71.4
Training in Supervision
No 11 52.4
Yes 10 47.6
If yes, source of training
On-the-job 2 20.0
Academic coursework 1 10.0
Other 1 10.0
More than one source 6 60.0
Sample Demographics
A full description of both counselor (n = 73) and su-
pervisor (n = 21) demographics is contained in Table
2. In general, the average counselor in this study was
44 years old and was more likely to be a nondisabled,
White female. She had approximately 10 years expe-
rience working as a state vocational rehabilitation
counselor and most likely did not retain any profes-
sional certification or licensure. Educationally, most
(71%) counselors had a master’s degree with fewer
than half in rehabilitation counseling. The average
caseload size for counselors was 127 clients. Super-
visors, as a group, were older (M = 50.33) and also
tended to be persons who were nondisabled, White
females. Supervisors had worked, on average, about
14 years as state vocational rehabilitation counselors
and, in addition, over 8 years experience as a VR su-
pervisor. The majority (71%) of supervisors had no
professional certification or license but most (81%)
had received a master’s degree with half being in re-
habilitation counseling. Each supervisor was respon-
sible for supervising about 6 counselors and, in addi-
tion, two-thirds of supervisors also carried a client
caseload of 32 persons as well. In terms of training in
clinical supervision about one-half of supervisors in-
dicated that they had received training prior to the in-
tervention as part of this study.
Results
Assessment of Training Impact
Supervisors. In terms of the four outcome measures
reported by VR supervisors, only changes in clinical
supervision knowledge were found. Specifically,
tests to assess changes between baseline to posttest 1
F(1,18) = 11.36, p = .003, d= .39 from posttest 1 to
posttest 2 F(1,11) = 7.48, p = .019, d= .41, baseline
to posttest 2 F(1,12) = 15.09, p = .002, d= .56 and
across all three observations from baseline (M =
99.75) to posttest 1 (M = 106.83) to posttest 2 (M =
112.08) F(1,11) = 12.40, p = .005, d= .53 were
found. As noted in the reported effect sizes (d), using
Cohen’s (1992) guideline, effect changes between
observations were between the “small” (.30) and
“medium” (.50) range. None of the other three out-
come measures (clinical supervision behavior, effi-
cacy and working alliance) across observations
(baseline to posttest 1, posttest 1 to posttest 2, base-
line to posttest 2 and baseline to posttest 1 to posttest
2) resulted in statistically significant differences. As
noted in Table 3, means across the other three out-
come variables and observation periods were similar
to one another.
As far as behavioral indices, on average, supervisors
reported that, within a 30-day period, they provided
an average of 19.2 individual supervision sessions at
baseline. This average, however, was inflated due to
one supervisor who reportedly met on 240 occasions
during the one-month period. If this one outlier were
removed, the average would be 7.6. Subsequent
post-test assessments taken six weeks later indicated
an average of 7.22 individual supervision meetings
but this level decreased six months later at the second
follow-up assessment where supervisors met, on av-
erage 5.15 occasions during the prior 30-day period.
If a supervisor met at least one or more times for indi-
vidual supervision, with each observation, the aver-
age individual supervision session lasted approxi-
mately 23 minutes. Although no significant change
was observed six weeks after the initial training,
there was a decrease by 3 minutes on average by the
end of the 6-month training period (see Table 3).
The average frequency of group supervision at base-
line was nearly 7 sessions per month but again, if one
were to remove the one outlier, a more accurate rep-
resentation at baseline would be .68 sessions during
the 30-day period. Subsequent observations at both
posttests six weeks after the training and six months
later indicated less than one group supervision ses-
sion that lasted about an hour per month.
Finally, in terms of how clinical supervision was ini-
tiated and consistent with counselor perceptions,
there was little change across observation periods, as
two-thirds of supervisors reported that counselors
were more likely to initiate supervision than supervi-
sors. Also consistent with counselor perceptions is
that when considering overall quality of supervision
(both individual and group), supervisors perceive
their supervision as between “somewhat valuable”
and “valuable” throughout the six-month period.
When asked how their supervision contributed to the
overall professional development of their counselors,
their perception throughout all three observation pe-
riods was that it made a “valuable” contribution (see
Table 3).
Counselors. Prior to reviewing results to assess
training impact, as noted in Table 4, with each suc-
cessive administration of the outcome measure-
ments, approximately one-half of the counselor sam-
ple elected not to continue. Thus, at baseline, there
were 73 respondents, at Posttest 1 (six weeks later)
there were 37 participants and, at Posttest 2 (5
Journal of Rehabilitation Administration, 38(1), 2014
Page 26 Herbert et al.
months later) there were 24 participants. As a result,
in order to assess training impact, it was decided to
rely on listwise comparison (i.e., inclusion of respon-
dents with complete data set of all outcome mea-
sures). This method, while reducing statistical power,
is often preferable to other methods for handling
missing data (Allison, 2001).
Mean changes across the three observation periods
on each outcome variables, resulted in no statistically
significant change between baseline, posttest 1 and
posttest observations. Thus, counselors did not per-
ceive any significant change in terms of clinical su-
pervision behavior, efficacy, knowledge and work-
ing alliance with their supervisors. Consistent with
these self-report outcome measures other behavioral
Journal of Rehabilitation Administration, 38(1), 2014
Clinical Supervision Training Page 27
Table 3
Clinical Supervision Outcome Variables Reported by Supervisors
Time of Behavior Efficacy Knowledge Working Alliance
Observation Mean SD Mean SD Mean SD Mean SD
Baseline 134.25 17.85 1088.00 222.12 99.75 13.76 128.83 16.84
Posttest 1 136.33 17.25 1178.25 158.85 106.83 10.46 128.33 12.32
Posttest 2 141.08 19.68 1184.67 187.96 112.08 12.99 128.67 18.33
n=12
Frequency of Individual Supervision Length of Individual Supervision
Mean SD Range n Mean SD Range n
Baseline 19.20 52.60 0-240 20 18.33 8.29 5-90 9
Posttest 1 7.22 9.20 1-30 18 23.33 12.24 10-60 9
Posttest 2 5.15 6.01 1-20 18 20.59 17.76 30-60 9
Frequency of Group Supervision Length of Group Supervision
Mean SD Range n Mean SD Range n
Baseline 12.70 37.72 0-240 10 47.86 37.06 10-120 7
Posttest 1 .40 .97 1-30 10 69.56 49.73 1-120 9
Posttest 2 .70 .95 1-20 10 57.00 40.87 10-120 5
Arrangement Method of Supervision Sessions as Reported by Supervisors
Pre-arranged Supervisor Initiated Counselor Initiated
Frequency Percent Frequency Percent Frequency Percent
Baseline (n=20) 5 25.0 2 10.0 13 65.0
Posttest 1 (n=18) 4 22.2 2 11.1 12 66.7
Posttest 2 (n=13) 3 23.1 2 15.4 8 61.5
Level of Satisfaction of Contributed to Counselor to
Clinical Supervision Provided Professional Development
Mean SD Range Mean SD Range
Baseline (n=21) 2.10 .89 1-5 2.33 .80 1-4
Posttest 1 (n=18) 2.72 1.45 1-5 2.11 .83 1-4
Posttest 2 (n=13) 2.58 1.16 2-5 2.33 .78 1-4
indicators including frequency and duration of indi-
vidual and group supervision also resulted in no sta-
tistically significant change across baseline, initial
post-test (six weeks after the on-site training pro-
gram) and subsequent post-test (six months of fol-
low-along consultation). As noted in Table 4, coun-
selors reported meeting for individual supervision at
a level almost identical at baseline and six months
later after supervisors received training in clinical su-
pervision. In short, counselors reported meeting on
average about five times per month. While individual
supervision increased by six minutes from baseline,
at the end of the training period six months later, it
actually decreased to a level lower than assessed at
baseline. Listwise comparisons of the 18 counselors
who completed all three observations indicated no sta-
tistical differences in terms of frequency of individual
supervision F(1, 17) = .007, p = .93.
Although a similar decrease in the frequency of pro-
viding group supervision occurred over time, this may
have largely been attributable to an outlier at baseline
and, if this individual was removed from the analysis,
it appears that counselors received less than one group
supervision session each month. The length of the ses-
sion over the course of the six-month period increased
from an hour at baseline to an hour and 40 minutes at
posttest 2.
Finally, in terms of how clinical supervision was initi-
ated, there was little change across observation peri-
Journal of Rehabilitation Administration, 38(1), 2014
Page 28 Herbert et al.
Table 4
Clinical Supervision Outcome Variables Reported by Supervisors
Time of Behavior Efficacy Knowledge Working Alliance
Observation Mean SD Mean SD Mean SD Mean SD
Baseline 118.93 30.89 1026.01 399.52 100.74 22.72 102.93 25.88
Posttest 1 115.81 31.39 955.76 413.47 97.70 22.96 96.78 27.85
Posttest 2 127.63 30.06 1135.33 272.13 104.83 19.07 109.83 22.90
Frequency of Individual Supervision Length of Individual Supervision
Mean SD Range n Mean SD Range n
5.06 5.37 0-30 68 20.5 19.23 0-120 62
3.78 4.29 0-20 36 26.62 19.48 5-60 32
4.79 6.19 0-30 24 17.75 9.10 5-30 20
Frequency of Group Supervision Length of Group Supervision
Mean SD Range n Mean SD Range n
2.64 14.39 0-120 73 64.41 51.81 1-180 29
.78 1.55 0-8 36 56.47 41.08 12-180 14
.46 .59 0-2 24 101.50 80.14 0-240 10
Arrangement Method of Individual Supervision Sessions as Reported by Counselors
Pre-arranged Counselor Initiated Supervisor Initiated
Frequency Percent Frequency Percent Frequency Percent
Baseline 9 12.3 47 64.4 17 23.3
Posttest 1 7 19.4 19 52.8 10 27.8
Posttest 2 5 20.8 15 62.5 4 16.7
Note. Baseline (n=73); Posttest 1 (n=37); Posttest 2 (n=24)
ods, as nearly two-thirds of counselors reported that
they were more likely to initiate supervision than
their supervisors. Still, when evaluating the overall
quality of supervision (both individual and group),
counselors rated their supervision received between
being “somewhat valuable” and “valuable” through-
out the six month period. There was also a minority
of counselors who perceived the quality of supervi-
sion to be less than desirable and perceived it as
“minimally valuable”, “not valuable” or “counter-
productive” (21.9% at baseline; 19.5% at post-test 1
and 16.7% at post-test 2).
Discussion
In accounting for why there were no observed
changes on all but one outcome measure as reported
by counselors and supervisors, several explanations
are offered. First, and perhaps most obvious, is that
the training intervention simply was not effective.
Although an increase in clinical supervision knowl-
edge was observed by supervisors throughout the
training program, as far as other impact areas, it did
not result in promoting more effective clinical super-
vision behavior, increasing one’s belief to conduct
clinical supervision, or enhancing the counselor-su-
pervisor relationship as it applied to clinical supervi-
sion practices. In terms of how counselors viewed
these changes, they too, failed to observe any
changes during the data collection period including
whether individual supervisors had increased their
knowledge pertaining to clinical supervision. Given
high inter-correlations among the four outcome mea-
sures which ranged from .73 (supervision knowledge
and working alliance [supervisor form]) to .88 (su-
pervision self-efficacy and working alliance [coun-
selor form]) (see Table 1), it follows that lack of
change in one area would be associated with a lack of
change in other areas. Still, one could raise the ques-
tion, “If change occurred in supervision knowledge,
then why not in the other domains?”
From an outsider perspective assessed by the VR
counselor, evaluating supervisor knowledge of clini-
cal supervision is something that can only be inferred
in terms of demonstrating this knowledge through
action or behavior. Consequently, even though su-
pervisors may have reported a continuing increase in
knowledge, unless it can be translated and experi-
enced through supervisory actions within the super-
vision relationship, it is difficult to infer from a coun-
selor perspective. Given that counselors did not re-
port any perceived changes in supervision in any
domain (knowledge, behavior, self-efficacy, and
working alliance) and the statistically significant and
high correlations that existed among outcome mea-
sures, it is not surprising that observed change did not
occur in other domains. Although this explanation
may account for the lack of change from a counselor
perspective, using the same logic, it does not follow
for why supervisors would report changes in clinical
supervision knowledge but not other behavioral,
self-efficacy, and working alliance domains. In ac-
counting for this explanation, it would seem that
training intervention as implemented had a differen-
tial impact only on the knowledge domain—but
why?
An immediate analysis of the 3-day on-site training
intervention indicated that the 22 supervisors who at-
tended the session (one person did not complete the
baseline assessment) provided a favorable review re-
garding presentation content (e.g., usefulness of
ideas, skills, and techniques, effectiveness of visual
aids) and presenter characteristics (e.g., “ability to re-
late to the group,” “effective response to questions,”
and “expertise on topic”). Specifically, an evaluation
scale completed anonymously by each supervisor
showed that, on average, participants perceived the
training program as “excellent” (overall average
3.8/4.0 scale). Additional narrative comments re-
garding the value of the on-site training program and
its impact on how participants would conduct super-
vision in the future also substantiated this impres-
sion. Thus, on the basis of these self-report com-
ments, it would seem that supervisors clearly valued
the training program and, in particular, increased
their knowledge about clinical supervision.
More developed aspects that require changes in per-
ceptions regarding one’s belief to implement effec-
tive clinical supervision practices (self-efficacy),
demonstrating this knowledge into effective clinical
supervision actions (behavior) and implementing it
within the counselor-supervisor relationship (work-
ing alliance) is something that develops over time,
however. In fact, this was a major reason why the
training intervention included a 6-month fol-
low-along training program and, in essence, provided
opportunity to increase confidence that supervisors
could provide clinical supervision, and, in doing so,
result in an increase effective supervisory behavior
and ultimately enhance the clinical supervision rela-
tionship between counselor and supervisor.
Despite any potential gains that may have been real-
ized from the initial training, outcome data clearly in-
dicated that it was not carried forward as part of the
Journal of Rehabilitation Administration, 38(1), 2014
Clinical Supervision Training Page 29
6-month follow-along training program. Perhaps
some of the early momentum was lost due to the
scheduling problems of the follow-along program
which began during the winter holiday season. Still,
the problem of consistent attendance was something
that continued throughout the entire follow-along pe-
riod. In fact, participation rates were less than 25%
during the initial two scheduled sessions. An exami-
nation of the overall participation rate across the
six-month follow-along program was, on average,
fewer than four sessions per supervisor of a possible
12-session program. A review of participation rates
indicated that of the 22 supervisors who completed
the on-site program, 13 participated in three or fewer
sessions (two persons did not participate in any ses-
sions), which represented 60% of the sample. The
highest number of attended sessions by any supervi-
sor was nine, which only one person achieved. Para-
doxically, a comment sometimes expressed by sev-
eral supervisors during follow-along sessions was
“how valuable it was to listen to other supervisors” or
that it “helped me in dealing with a similar situation
that I have with one of my counselors.” Despite this
feedback, for whatever reason, most supervisors did
not make the follow-along training a priority as part
of the other duties. During the last supervision ses-
sion, supervisors were asked directly about their at-
tendance to better understand why most supervisors
were not more consistent. Although the obvious so-
cial desirability effect cannot be ignored, information
learned from this discussion indicated that supervi-
sors had (a) other pressing concerns to attend to and
simply could not make the commitment needed, (b)
not fully understood the commitment they were tak-
ing on in terms devoting approximately 75 minutes
every two weeks, (c) unexpected schedule conflicts,
and, in a few instances, (d) problems connecting to
the teleconference call system.
To some extent, the issue of competing work priori-
ties and participating in an on-going 75-minute group
consultation session every two weeks constituted a
commitment that many supervisors were not willing
to make. For example, two-thirds of participating su-
pervisors carry their own client caseloads in addition
to supervising counselors. In fact, four supervisors
reported caseloads of 60 or more clients (range was 3
to 130 with an average of 32 clients). This situation
was necessary because of counselors who left that
agency and, as a result, supervisors were required to
“pick up the slack.” Another clear demand on work-
load was the number of counselors required to super-
vise. Given a reported average supervision load of
nearly 6 counselors per supervisor (range 3 to 7 coun-
selors), each of which had an average caseload of
130, one can quickly see that there would be strong
demands placed on supervisor time and effort in
monitoring this workload. Thus, when supervisors
report that there are many pressing demands on their
time, there is evidence to support this belief.
Another speculative account for why there was not
greater participation in the follow-along portion of
the training program may also be partially explain-
able due to the centrality of counseling as part of
one’s professional identity. For example, less than
one-third of supervisors were Certified Rehabilita-
tion Counselors (15% of counselors were CRCs)
and, of those with master’s degrees, less than
one-half were in rehabilitation counseling. One in
five supervisors with graduate degrees had them in
“other” areas including human resources develop-
ment, management, special education, psychology,
or other counseling subspecialty (e.g., substance
abuse), and another 20% had baccalaureate degrees.
Further support for questioning the centrality of
counseling for both VR counselors and supervisors
are indicative in several written comments shared as
part of the on-line survey. Two examples from coun-
selor comments noted on the on-line survey may
highlight this problem:
I think so much emphasis is placed on policies
and procedures [that] supervisors and counsel-
ors tend to forget the importance of our counsel-
ing skills and techniques. I think it will be good
practice for supervisors to remind and consult
counselors on counseling skills and techniques
so quality services can be provided.
My agency is much more interested in managing
activity due lists, saving money, and producing
good statistics than they are in the issues that you
have presented, that is, providing good counsel-
ing to consumers, improving counseling skills.
Providing good counseling is a strong value for
me and I often feel out of sync with the priorities
of the agency. I spend 3/4 of my time doing ad-
ministrative tasks. The emphasis of my supervi-
sor’s direction has been for me to increase my
time for administrative tasks, take short cuts in
working with consumers and to not provide
counseling outside of initial interviews, plan de-
velopment and updates. Increasing counseling
skills is not valued.
One other supplemental observation related the cen-
trality of counseling as part of state VR practice is
Journal of Rehabilitation Administration, 38(1), 2014
Page 30 Herbert et al.
that on several occasions during the initial 3-day
training, it was clear that some supervisors lacked a
clearly developed counseling orientation. For exam-
ple, when asked as part of the initial on-site training
program, “How do you assess your supervisee’s
counseling orientation?” two themes emerged. First,
few supervisors could articulate their own counseling
orientation and second, even fewer understood the
relevance of asking this question in terms of its impli-
cation to rehabilitation practice. Parenthetically,
there were some supervisors who, although later ac-
knowledging the importance of understanding a
counselor’s theoretical orientation as it serves as a
framework that explains beliefs about facilitating cli-
ent change and, within that perspective, a rationale
for why counselors behave the way they do, ex-
pressed that knowing such was only appropriate for
counselors-in-training and not current counselors. If
limited discussion as part of supervision fails to ad-
dress questions about one’s counseling orientation as
part of one’s own experience as a counselor, it seems
reasonable that subsequent discussion about this
topic will not occur when they are promoted as super-
visors. As a result, greater focus and comfort with ad-
dressing administrative rather than clinical supervi-
sion aspects seems more likely with supervisors who
do not possess or emphasize a counseling focus.
Limitations
A major limitation in the study and one often found
with repeated measures designs concerns participant
attrition (Van Horn, Green & Martinuseen, 2009).
Despite multiple email notifications to encourage
continued participation throughout the study, these
efforts were unsuccessful. Whether compensating
participants as originally planned in the research de-
sign would have maintained involvement is conjec-
ture. Outside of receiving approved continuing edu-
cation units for persons who were also Certified Re-
habilitation Counselors, there were no other
incentives allowable. Unfortunately, only 15% of
counselors and 29% of supervisors (at baseline) were
also Certified Rehabilitation Counselors.
The study used a convenience sample of counselors
and supervisors who perhaps may have had some
predilection for clinical supervision. One way to ad-
dress this problem would be to use random assign-
ment but, given the number of available supervisors
to participate in the study, separating supervisors into
two groups would not have produced the statistical
power needed to test the major research questions of
interest as they pertained to supervisors (i.e., Did
change occur across three observations on the four
outcome measures?). For instance, in an a priori use
of a software program, G*power (Faul & Erdfelder,
1992), we used a conservative estimate of statistical
power (.80) in order to evaluate whether differences
would occur. Although somewhat arbitrary, this esti-
mate level is one often used in social sciences re-
search (Cohen & Cohen, 1983). We also used a con-
servative Type I error (.05) estimate in order to detect
small effect size differences (.25) across three obser-
vations. Given that we had no prior studies to evalu-
ate effect sizes, we again employed a conservative
approach to evaluate small effect sizes that might oc-
cur. Consequently, not wanting to sacrifice statistical
power for the benefit of randomization, we decided
to follow a more robust approach to address the re-
search questions. The limitation of doing so was that
confidence in our results in generalizing to the popu-
lation of state VR supervisors may be somewhat
compromised.
Future Research
The potential impact of this training intervention
cannot be fully evaluated unless the program, as con-
ceived, is actually implemented. Thus, additional
study with other state VR samples are needed that
may more readily follow the protocol as outlined.
Further, having greater numbers of supervisors
would allow for randomized control trials to occur
whereby any bias for wanting to participate in the
training may be mitigated somewhat. Given that par-
ticipation is voluntary, it may be that even with ran-
domization, one could not truly rule out the problem
of self-selection bias. Clearly, there may be differ-
ences between supervisors who elect and do not elect
to participate and, should changes in outcomes result,
one could not rule out the possibility that any ob-
served differences may be attributable to individual
differences rather than the training program per se.
Multiple applications with various SVR agencies us-
ing the standardized training content program could
provide a better indication of generalizability.
Understanding how clinical supervision impacts
counselor behavior and ultimately client vocational
rehabilitation outcome is perhaps the most important
reason for continuing this research. However, until
an effective intervention can be developed and vali-
dated, there is no “road map” that can assist supervi-
sors to enhance their clinical supervision competen-
cies. Previous studies (e.g., Herbert, 2004a; Herbert
& Trusty, 2006; Schultz et al., 2002) as well as the
current investigation continue to support that both
Journal of Rehabilitation Administration, 38(1), 2014
Clinical Supervision Training Page 31
counselors and supervisors are basically satisfied
with how supervision is conducted or perhaps more
accurately stated, not conducted. Until evidence
demonstrates that clinical supervision can impact
counselor behavior and ultimately client vocational
rehabilitation outcomes, there seems to be little rea-
son from a systems perspective for changing current
practice. It is interesting to note, however, a study by
McCarthy (2013) that examined the supervisory
working alliance as perceived by state/federal voca-
tional rehabilitation counselors and its impact on cli-
ent vocational rehabilitation outcomes. Results of
155 VR counselors from five states suggested that,
although the supervisory working alliance did not
predict successful VR outcomes, it had a differential
impact with clients assigned to counselors with two
or less year work experience. Accordingly, with each
unit increase in the working alliance measure, three
additional case closures resulted each year. McCar-
thy also noted that these counselors (44% of whom
where CRCs) reported spending 69 minutes per week
in clinical supervision. In comparison to our study,
15% of counselors were CRCs who reported devot-
ing about an hour each month over the six-month
training program. The McCarthy study suggests that
more intensive clinical supervision has the potential
to contribute to a good working alliance particularly
among newer counselors and, as a result, produce
greater numbers of successful VR closures for clients
assigned to work with newer counselors. Continued
work is needed not only to better understand which
supervisor behaviors are related to outcomes but, as it
relates to the present study, whether training can suc-
cessfully produce change in supervision practices for
both newer and more experienced counselors. The
relationship of being a Certified Rehabilitation
Counselor and its impact on clinical supervision
from both counselor and supervisor perspectives also
merits further attention.
References
Allison, P. D. (2001). Missing data. Sage University
papers series on quantitative applications in so-
cial sciences, 7-136. Thousand Oaks, CA: Sage.
Bandura, A. (1997). Self-efficacy: The exercise of
control. New York, NY: W. H. Freeman & Com-
pany.
Bandura, A. (2006). Guide for constructing self-ef-
ficacy scales. In F. Pajares & T. Urdan (Eds.),
Self-efficacy beliefs of adolescents (pp.
307-337). Greenwich, CT: Information Age.
Bernard, J. M., & Goodyear, R. K. (2004). Funda-
mentals of clinical supervision (3rd ed.). Boston,
MA: Pearson Education.
Borders, D. L. (1989). A pragmatic agenda for de-
velopmental supervision research. Counselor
Education & Supervision, 29, 16–24.
Bozarth, J. D., Rubin, S. E., Krauft, C. C., Bolton, B.
& Richardson, B. K. (1975). Client counselor in-
teractions: Patterns of service: Client outcome:
Overview of project, conclusions and implica-
tions: Arkansas Studies in Vocational Rehabili-
tation Series Monograph 28. University of
Arkansas, Arkansas Rehabilitation Research and
Training Center in Vocational Rehabilitation AR.
Retrieved from www.naric.com (Document
#O07069).
Capella-McDonnall, M. E. (2005). Predictors of
competitive employment for blind and visually
impaired consumers of vocational rehabilitation
services. Journal of Visual Impairment & Blind-
ness,99, 303–315. Retrieved from www.naric.
com (Document #J49162).
Cohen, J. (1992). A power primer. Psychological
Bulletin, 112, 155–159.
Cohen, J., & Cohen, P. (1983). Applied multiple re-
gression/correlation analysis for the behavioral
sciences (2nd ed.). Hillside, NJ: Lawrence Erlbaum
and Associates.
Donnell, C. M., Lustig, D. C., & Strauser, D. R.
(2004). The working alliance: Rehabilitation
outcomes for persons with severe mental illness.
Journal of Rehabilitation, 70, 12–18. Retrieved
from www.naric.com (Document #J47599).
Efstation, J. F., Patton, M. J., & Kardash, C. M.
(1990). Measuring the working alliance in coun-
selor supervision. Journal of Counseling Psy-
chology, 37, 322–329.
English, W. R., Oberle, J. B., & Byrne, A. R. (1979).
Rehabilitation counselor supervision: A national
perspective [Special Issue]. Rehabilitation
Counseling Bulletin, 22, 7–123.
Faul, F., & Erdfelder, E. (1992). G*POWER: A pri-
ori, post-hoc, and compromise power analysis
for MS DOS [Computer program]: Bonn, FRG:
Bonn University, Department of Psychology.
Garske, G. G., & Soriano, M. (1997). Client percep-
tions of the rehabilitation counseling relation-
ship: Humanistic approaches and related
outcomes. Journal of Applied Rehabilitation
Counseling, 28(2), 10–14.
Journal of Rehabilitation Administration, 38(1), 2014
Page 32 Herbert et al.
Herbert, J. T. (1995). Clinical supervision. In A. E.
Del Orto, & R. P. Marinelli (Eds.), Encyclopedia
of disability and rehabilitation (pp. 178–190).
New York, NY: Macmillan.
Herbert, J. T. (2003, March). Clinical Supervision
within the state-federal vocational rehabilitation
program: A combined qualitative and quantita-
tive analysis. U. S. Department of Education,
National Institute on Disability and Rehabilita-
tion Research, Mary E. Switzer Scholarship
Meeting, Washington, DC.
Herbert, J. T. (2004a). Qualitative analysis of clini-
cal supervision within the public vocational re-
habilitation program. Journal of Rehabilitation
Administration,28, 51–74.
Herbert, J. T. (2004b). Analysis of clinical supervi-
sion practices as documented in rehabilitation
counseling syllabi and fieldwork manuals. Reha-
bilitation Education, 18, 13–33.
Herbert, J. T. (2004c). Clinical supervision. In D. R.
Maki & T. F. Riggar (Eds.), Handbook of Reha-
bilitation Counseling (2nd ed., pp. 305–317).
New York, NY: Springer.
Herbert, J. T. (2009, October). Clinical supervision
of master’s level state VR counselors: A training
model. Presentation given to the National Coun-
cil on Rehabilitation Education, Rehabilitation
Services Administration, and Council of State
Administrators of Vocational Rehabilitation Na-
tional Training Conference on Rehabilitation
Education, Washington, DC.
Herbert, J. T., & Caldwell, T. (in press). Clinical su-
pervision in rehabilitation counseling practice.
In F. Chan, N. L. Berven, & K. R. Thomas (Eds.),
Counseling theories and techniques for rehabili-
tation health professionals. New York, NY:
Springer.
Herbert, J. T., & Trusty, J. (2006). Clinical supervi-
sion practices and satisfaction within the public
vocational rehabilitation program. Rehabilita-
tion Counseling Bulletin,49, 66–80.
Kissinger, D. B. (2008). Traumatic brain injury and
employment outcomes: Integration of the work-
ing alliance model. Work: A Journal of Preven-
tion, Assessment, and Rehabilitation,31,
309–317. Retrieved from www.naric.com (Doc-
ument #J55565).
Lustig, D. C., & Strauser, D. R. (2009). Rehabilita-
tion counseling graduate students’ preferences
for employment: Agreement between actual and
perceived job tasks of state-federal vocational
rehabilitation counselors. Rehabilitation Coun-
seling Bulletin, 52, 179–188.
Lustig, D. C., Strauser, D. R., Rice, D., & Rucker, T.
(2002). The relationship between working alli-
ance and rehabilitation outcomes. Rehabilitation
Counseling Bulletin, 46, 25–33.
McCarthy, A. K. (2013). Relationship between su-
pervisory working alliance and client outcome
between state vocational rehabilitation counsel-
ing. Rehabilitation Counseling Bulletin, 57,
23–30.
Olshansky, S. (1973, January-February). Six scan-
dals in vocational rehabilitation. Rehabilitation
Record, 23–35. Retrieved from www.naric.com
(Document #J03280).
Saunders, J. L., Leahy, M. J., McGlynn, C., &
Estrada-Hernandez, N. (2006). Predictors of em-
ployment outcomes for persons with disabilities:
An integrative review of potential evi-
denced-based factors. Journal of Applied Reha-
bilitation Counseling, 37(2), 3–20.
Schultz, J. C. (2008). The tripartite model of super-
vision for rehabilitation counselors. Journal of
Applied Rehabilitation Counseling, 39(1),
36–41.
Schultz, J. C., Ososkie, J. N., Fried, J. H., Nelson, R.
E., & Bardos, A. N. (2002). Clinical supervision
in public rehabilitation counseling settings. Re-
habilitation Counseling Bulletin, 45, 213–222.
Stebnicki, M. A., Allen, H. A., & Janikowski, T. P.
(1997). Development of an instrument to assess
perceived helpfulness of clinical supervisory be-
haviors. Rehabilitation Education, 11, 307–322.
Stenack, R. J., & Dye, H. A. (1982). Behavioral de-
scriptions of counseling supervision roles. Coun-
selor Education and Supervision, 21, 295–304.
Strauser, D. R., Lustig, D. C., & Donnell, C. (2004).
The relationship between working alliance and
therapeutic outcomes for individuals with mild
retardation. Rehabilitation Counseling Bulletin,
47, 215–223. Retrieved from www.naric.com
(Document #J48148).
Tabachnick, B. G., & Fidell, L. S. (2001). Using
multivariate statistics (4th ed.). Boston, MA:
Allyn and Bacon.
Theilsen, V. A., & Leahy, M. J. (2001). Essential
knowledge and skills for effective clinical super-
vision in rehabilitation counseling. Rehabilita-
tion Counseling Bulletin, 44, 196–208.
Journal of Rehabilitation Administration, 38(1), 2014
Clinical Supervision Training Page 33
Thomas, D. M., & Whitney-Thomas, J. (1996). Per-
spectives from consumers and counselors on ele-
ments that influence successful vocational
rehabilitation system delivery. Research to
Practice,2, 1–2. Retrieved from www.naric.com
(Document #O13211).
Van Horn, P., Green, K., & Martinuseen, M. (2009).
Survey response rates and survey administration
in counseling and clinical psychology: A
meta-analysis. Educational & Psychological
Measurement, 69, 389–403.
Worthington, E. L., & Roehlke, H. J. (1979). Effec-
tive supervision as perceived by beginning coun-
selors-in-training. Journal of Counseling
Psychology, 26, 64–73.
Author Notes
James T. Herbert, PhD, is a professor and Program
Coordinator for the Rehabilitation and Human Ser-
vices at Penn State University.
Soo-yong Byun, PhD, is an assistant professor in
Educational Theory and Policy at Penn State.
Jared C. Schultz, Ph.D. is a professor and Director
of the Rehabilitation Counseling Program at Utah
State University.
Meritza Tamez, PhD, recently completed her doc-
toral degree in Counselor Education and Supervision
at Penn State University.
Heather A. Atkinson, MA, is a doctoral student in
Counselor Education and Supervision at Penn State
University.
Funding for this study was provided through a Na-
tional Institute on Disability and Rehabilitation Re-
search grant (Mary E. Switzer Research Fellowship
Award #H133F100003, Dr. Leslie Caplan, Project
Officer). The contents expressed in this paper do not
necessarily represent the policy of the U.S. Depart-
ment of Education and should not assume endorse-
ment by the federal government.
Further information regarding this study should be
directed to the Principal Investigator, James T. Her-
bert at The Pennsylvania State University, Depart-
ment of Educational Psychology, Counseling, and
Special Education, 314 CEDAR, University Park,
PA, 16802; Email: jth4@psu.edu. This project is
dedicated to the memory of Roger Barton, an exem-
plary advocate for persons with disabilities who had
a distinguished career with the Pennsylvania Office
of Vocational Rehabilitation and who served as a
member on our Research Advisory Board.
Journal of Rehabilitation Administration, 38(1), 2014
Page 34 Herbert et al.
