Follow-up art therapy study 1
Running head: FOLLOW-UP STUDY ON ART THERAPY WITH PRISON INMATES
Effects of Art Therapy with Prison Inmates:
A Follow-up Study
David Gussak
The Florida State University
(images not attached)
Follow-up art therapy study 2
Effects of Art Therapy with Prison Inmates:
A Follow-up Study
In the summer of 2003, a pilot study was conducted to quantify the effects of art therapy with
prison inmates (Gussak, 2004). This study yielded data that supported the positive effects of art
therapy in prison (Gussak, 2004). This quasi-experimental study utilized a standardized art therapy
assessment, the Formal Elements Art Therapy Scale (FEATS) (Gantt & Tabone, 1998), and a preand post- test survey designed by the primary researcher and the mental health counselor of the
prison unit to measure the effects. The results indicated that over four weeks, two sessions a week of
art therapy groups, the inmates who participated demonstrated a significant decrease in depressive
symptoms and improvements in mood. Results based on the data indicated that the participants’
attitudes and acceptance of each other and the environment also improved, and interaction improved
with peers and staff. Overall, those who participated demonstrated greater compliance with
directives and improved behavior. These results encouraged a quantitative follow-up study
conducted in the summer of 2004. This follow-up study will report presents the methodology used
including the FEATS and the Beck Depression Inventory-Short Form as pre and post-test
assessments, the results and future plans for art therapy in the prison system. Finally, a case vignette
is presented as an example for the efficacy of alleviating depression in art therapy in prison.
Literature Review
Based on the findings of the pilot study, this study focused primarily on alleviating
depression in prison inmates. Depression is a major problem in prison. In 1994, Eyestone and
Howell found that of 102 inmates evaluated at a variety of different security levels, 25%
maintained symptoms of severe depression As well, another 30% maintained depressive-like
symptoms, but did not meet all of the standards of the DSM-III-R (1987) classification system
Follow-up art therapy study 3
used. Chiles, von Cleve, Jemelka and Turpin (1990) found that 30% of all male inmates
maintained a major depression diagnosis while 15% were diagnosed with dysthymia.
Some maintain that depression is a normal reaction to incarceration. However, long-term,
depressed inmates “can affect the prison community as a whole” (Boothby & Durham, 1999, p.
110). Depressed inmates may suicide, or self-abuse (Toch, 1992), which is, of course,
detrimental to the inmates, and can be quite costly to the prison system. As well, an increase of
depression in the prison population may increase drug and alcohol use as a means to escape
(Winfree, Mays, Crowley, & Peat, 1994). All of these reactions to depression cost the prison
system money and staff-time to monitor such reactions and for psychotropic medication
administered to alleviate these mental illnesses (Baillargeon, Black, Contreras, Grady &
Pulvino, 2002). Art therapy has been used as a viable, alternative means to address depression.
It is believed that the actual art making process helps alleviate depressive symptoms. The
art making process provides an alternate means of escape (Gussak, 1997, 2004). It can also instill
a sense of self worth and identity. However, there has only been the pilot study thus far that
attempted to assess empirically the benefits of art therapy in a prison population, including
alleviating depression.
Method
A control group pre-test/post-test design was implemented for this study. It was
hypothesized that if inmates receive art therapy services, then they will exhibit marked
improvement in mood, socialization and problem-solving abilities within the correctional
environment.
Follow-up art therapy study 4
Participants
This follow-up study was conducted in a medium to maximum-security male adult
correctional facility in a rural section in Florida’s panhandle. An announcement asking for
volunteers for this project was made on all of the units. Participants who volunteered for this
study were randomly assigned to either the control group or the experimental group. Twentyseven inmates were provided art therapy services. The age range for the participants was from 21
to 59 years old. Seventy-four percent of the participants were white, 22% were black and 4%
were Hispanic. Forty-four percent of them had completed high school or an equivalent (GED)
while 22% had attended some college. For 41% of the participants, this was their first prison
sentence. Their crimes ranged from possession of narcotics to murder. Only two of the
participants were not taking medication for a mental illness. Of these 27 participants, only 16
attended all 8 sessions and completed the post-tests.
The control group was not provided art therapy sessions, but was administered the pre
and post assessments. Seventeen volunteers attended the initial session when the pre-test
assessments were administered. The age range for the control group was from 21 to 59 years old.
Eighty-two percent were white, 18% were black and 0% were Hispanic. Forty-seven percent of
them had completed high school or an equivalent (GED), while one member (5%) of the control
group had attended some college. For 23% of the participants, this was their first prison sentence.
Seventy-seven percent of the control group was taking medication for a mental illness. Of the 17
members of the control group, 13 completed all of the assessment procedures.
Measures
This research project used two measurement tools. Each client was asked to complete a
drawing using standardized materials of a person picking an apple from a tree (PPAT) the first
Follow-up art therapy study 5
day and the last day of sessions. These drawings were used for a pre-and post project
comparison, and were assessed using the Formal Elements Art Therapy Scale (FEATS) rating
guide (Gantt & Tabone, 1998). This assessment demonstrated much success in ascertaining a
decrease in depressive symptoms in the pilot study; thus, it was believed it would once again
prove a valuable tool in ascertaining change in mood in the follow-up study.
The FEATS is composed of a 14-point scale with a possible score of 0 through 5,
focusing on such characteristics as: prominence of color and color fit; implied energy; space;
integration; logic; realism; problem-solving; developmental level; details; line quality; person;
rotation and perseveration. The FEATS was designed primarily to assess the presence of four
major diagnoses: 1) major depression; 2) bipolar disorder, mania; 3) schizophrenia; and 4)
delirium, dementia, amnestic and other cognitive disorders. However, in the pilot study the
FEATS was used an outcome measure to ascertain change, specifically in mood:
According to Gantt and Tabone, the diagnostic categories are assessed based on
the ratings of a combination of several characteristics. For example, it is assumed
that low ratings in prominence of color, color fit, energy, space, realism, details
and person reflect major depression (1998, p.26). In this study, although changes
were evaluated taking these diagnostic criteria into consideration, this tool was
also used to measure other changes, such as general socialization skills and
attitude towards the participating inmates’ situation. It was also assumed,
although this has not been explored in previous studies, that certain characteristics
may also be used independently. For example, if the problem-solving scale
demonstrates improvement, then the inmates may have gained problem-solving
abilities over the course of treatment (Gussak, 2004, p.247-248).
Follow-up art therapy study 6
In this study, if changes occurred in the drawing scales that ascertained depression and problemsolving between pre and posttests, then it was assumed that change in the participants’ mood and
problem-solving abilities occurred.
The PPATS were assessed using the FEATS by the primary researcher and the research
assistant for this project. The primary researcher had been specifically trained to administer this
assessment, and the research assistant had been trained in her graduate program, with additional
information provided by the primary investigator. What is more, the FEATS manual greatly
simplified the process, which “gives researchers ready-to-use tools for studying any
population… [t]hey can use our manual for training raters… (Gantt & Tabone, 1998, xvi).
Unlike the pilot study, which used a self-developed pre and post-test survey, the
participants of this study were asked to complete the Beck Depression Inventory-Short Form
(BDI-II), administered by the unit psychologist before and after the treatment period. The BDI-II
is a standardized psychological assessment (Beck, Rial &Rickets, 1974; Beck & Steer, 1993),
used to ascertain the intensity of depression. It consists of a questionnaire of 21 groups of
statements. The responses are weighed with a score between 0-3, based on emotional content of
the response, 0 indicating a bright mood or lack of depressive feelings, 3 indicates highly
depressive reaction. For example, the first group of statements reads as: 1. Sadness: 0=”I do not
feel sad”; 1=”I feel sad much of the time”; 2=”I am sad all of the time”; and 3=”I am so sad or
unhappy I can’t stand it”. The scored responses from all 21 statements are added together for
one final score. Total scores can range from 0 to 63. Cut-off scores are recommended; a score of
0 to 9 is described as minimal depression, scores of 10 to 16 are considered mild depression,
scores of 10 to 16 are considered mild depression…17 to 29 are described as moderate, and
scores of 30 to 63 are described as severe depression (Beck & Steer, 1993; Boothby & Durham,
Follow-up art therapy study 7
1999). Previously, the BDI-II has been used successfully to evaluate depression in prisoners
(Boothby & Durham, 1999).
For this study, the BDI-II was not used to measure the intensity of depression of those
participating in the study. Rather, with the guidance of the facility’s psychologist, it was simply
used to measure outcome, to see if there was a greater decrease in scores for those that
participated in the art therapy sessions compared to those that did not. The pre and post scores
were compared to ascertain change.
Procedure
Art therapy group sessions were conducted on a prison mental health treatment unit two
days a week, two sessions a day, during an eight-week period. Although 27 inmates were divided
up into 4 separate groups, only 16 attended the groups consistently; these 16 attended all eight
sessions, including the final session. The call for volunteers was presented by the unit
psychologist and the art therapist; the attendees were volunteers only. Several participants who
were chosen had attended art therapy sessions provided during the initial research project
(Gussak, 2004). Each of the treatment groups met once a week. Thus, each group member was
seen 8 times during the course of the program. One clinician, an art therapist, led the groups.
Prior to the sessions, the unit psychologist administered the Beck Depression InventoryShort Form to each of the group attendees. The first day of the sessions, each participant was
asked to draw a picture of a person picking an apple from a tree, per the instructions established
by Gantt and Tabone (1998). The art therapist administering the sessions met once a week with
the faculty supervisor and primary investigator to discuss progress, responses and possible art
therapy directives. As before, the art therapy directives primarily relied on the direction of the
individual art therapy group sessions. However, the project developed over the 8 weeks from
Follow-up art therapy study 8
simple to complex, and from individual art tasks to more group-focused activities. For example,
the beginning of the sessions consisted of the members of the group creating an outside-inside
box. This project represented a very individualized self-symbol.
Later on, the group members were asked to do a “poetry pass”. This consisted of each
member writing a sentence on a piece of paper in a specific color. After everyone has done this,
the papers are passed to the left, and each member writes a sentence responding to the sentence
above. Everyone then folds the paper so that only the current sentence is exposed, and then the
papers are passed once again. This continues until everyone has the opportunity to contribute to
everyone’s paper. This usually resulted in humorous, if not meaningful poems. Each member
was then expected to create an image in reaction to the poem. This began the group facilitation
process, while still maintaining an individual reaction. Following sessions focused on rendering
self-symbols from Sculpy culminating in placing each sculpted piece on large mural paper to
create an environment that incorporated everyone’s piece. This completed the transition from
individual identity to a sense of belonging to a larger group.
After the eight weeks of art therapy sessions, each participant was asked to complete a
BDI-II post-test and another drawing of a person picking an apple from a tree. Examples of these
directives are presented in the case vignette of Dean.
Results
Pre- and Post BDI-II. Sixteen members of the experimental group and 13
members of the control group completed a pre and post BDI-II assessment. The change in BDI-II
scores from pretest to posttest (i.e., posttest score – pretest score) was calculated and the
differences were analyzed using independent-sample t tests to find differences between the
experimental and control groups. The results were t(26) = -2.58, p < .05. The experimental group
Follow-up art therapy study 9
(M = -7.81, SD =9.81) had significantly greater decrease from pretest to posttest than the control
group (M = 1.00 , SD =7.59). The effect size of .15 was calculated using Cohen’s d equation.
FEATS. Sixteen members of the experimental group and 13 members of the
control group completed a post FEATS drawing. The drawings were blindly rated using the
FEATS by two investigators. The inter-rater reliability was determined using the kappa statistic.
This formula resulted in a score of .87, indicating a robust inter-rater reliability.
The changes in FEATS scores from pretest to posttest (i.e., posttest score – pretest score)
were calculated for each of the 14 categories and the differences were separately analyzed using
independent-sample t tests to find differences between the experimental and control groups. The
only significant result was for Rotation, t(30) = 2.42, p < .05. The effect size of .88 was
calculated using Cohen’s d equation. The experimental group’s rotation (M = .40, SD = 1.05)
was greater than the control group’s rotation (M = -.42, SD = .67).
Overall, the hypothesis, that inmates who receive art therapy services will exhibit marked
improvement in mood was supported by the BDI-II results. The results from the FEATS,
however, did not support the hypothesis. There was no quantitative data that supported the
hypothesis that there would be change in socialization and problem-solving abilities. Thus,
although it seems as if the art therapy was effective with the experimental population, the results
are mixed.
Discussion
Despite the changes indicated in the pilot study, and anecdotal evidence and the BDI-II
score, which seems to support the hypothesis, overall the FEATS for the follow-up did not seem
to indicate much change. In the pilot study, changes in several scales reflected a decrease in
depressive symptoms. “Improvement in five… categories (prominence of color, color fit,
Follow-up art therapy study 10
energy, details of objects and environment and space) supports the conclusion that there was a
decrease in depressive symptoms and an elevation of mood” (Gussak, 2004, p.249). However,
the follow-up only reflected change in one of the categories, Rotation, an irrelevant scale for the
sake of this study.
Several factors may have contributed to the lack of significance in the FEATS. For one
thing, unbeknownst to the researchers, due to the confidentiality of the data sets and the different
group leader for this project, several of the experimental participants for this study had also
participated in the pilot study. It was not until the project had already proceeded that this was
brought to the researcher’s attention. Thus, they were already familiar with the drawing
procedure, as they had completed two PPATs already. Thus familiarity with the procedure my
have resulted in a practice effect. As it was, several of the participants joked about having to do
“yet another person and apple tree drawing,” and even had fun with the procedure (For example,
see Figure #1).
Insert Figure 1 about here
Another factor may have been that those who volunteered, and completed the entire
procedure, did so because of an overriding interest in art making. Some of them may have
already had some skills, which may have resulted in little change between the drawings;
developed schemas and style may have prevented much change between the two drawings.
Another factor may have been the low participatory numbers. Although there were some paired
images that show very little change (Figures 2a and 2b),
Insert Figures 2a and 2b about here
there were some pre and post images that reflected significant differences (Figure #3a and #3b).
Insert Figures 3a and 3b about here
Follow-up art therapy study 11
It seems that the measurement tool may not have been sensitive enough to measure minute
change in small groups.
Despite a lack of significant differences on the FEATS between the control and
experimental groups, the BDI-II and individual cases bore out that there was a change in mood
and socialization. Thus, the findings are somewhat inconclusive, but it could be assumed that
indeed the art therapy was beneficial, but that the FEATS was not a useful tool in this case to
measure these changes. This was reflected in the work of Dean, a participant in the study. His
case vignette is included to illustrate the benefits art therapy seemed to have on the clients
despite the lack of corroborating data from the FEATS.
Case Vignette- Dean
Dean, a 60-year-old Caucasian man, was serving a 17-year sentence for a parole
violation. His initial sentence resulted in a 30 year paroled sentence for sexual assault of a child.
Allegations of sexual assault posited by his ex-wife initiated his second offense. Although he
was found not guilty for this offense, this was still considered a parole violation, and he was
subsequently sent back to prison to complete the rest of his initial sentence. By the time the art
therapist began working with Dean, he had already served 10 of his 17 years.
The art therapist who provided the sessions reported that he initially attended sessions
with a “chip on his shoulder”, believing he was attending art lessons, and did not need therapy
nor that he even belonged in prison. He tended to test limits with the art therapist, even
indicating at one point that the art therapist was “in the field to frustrate the poor inmates who are
incapable of retaliation.” (Later, it was discovered by the art therapist that Dean was convicted
based on the strength of drawings done by the child in court and submitted by a forensic art
Follow-up art therapy study 12
therapist. This was not conveyed until after trust was established between Dean and the art
therapist.)
He leaned towards self-pity, believing nobody had problems like him. He admitted to
feeling depressed. He did not interact with anyone in the groups, and generally did not talk to the
others except in disparaging terms. He did have a proclivity for creating art, which was evident
by his winning first prize in a drawing contest conducted among three prisons. Figure #2a
(example provided earlier) is his initial FEATS drawing.
Initially, Dean reacted superficially to the sessions, choosing to keep to himself and not
interact with others. Despite materials provided, Dean attended sessions with his own pencils,
and insisted on using the highly structured materials. He generally felt that his images were
“substandard” and he would subsequently destroy them.
He participated in the “poetry pass” (Figure 4).
Insert Figure 4 about here
However, when asked to focus on his reaction to the poem in the follow-up drawing, he chose to
focus on one line he contributed (“A strong person can pick themselves up, drawing from inner
strength”) and assigned his own meaning. When asked not to use his pencils, he chose markers,
and was surprised there was no eraser. He voiced irritation at the therapist, and then rendered
figure 5.
Insert Figure 5 about here
As Dean reflected on his drawing, he stated that inner peace is one of his goals. The
drawing is of a barren landscape with a barn far off in the distance, almost unattainable. Dean did
not show up for the following session, but the self-symbol series seemed to be a turning point for
Dean.
Follow-up art therapy study 13
After introducing the concept of self-symbols, the therapist asked all the participants to
choose and render a symbol they felt represented their character. Dean chose the lion and drew a
lion’s head in colored pencil (Figure 6).
Insert Figure 6 about here
He voiced contentment with this image, and remained quiet throughout the session. When asked,
he indicated that the symbol came from his past focus on inner strength.
After the session, he apologized to the therapist for missing the previous session, and then
began to explain to her why he was in prison, and how upset he was for not being able to see his
daughter. He began to tear up and admitted that he was quite angry and was very sad for losing
his “baby girl.” It was then that he expressed anger because forensic art therapy was used against
him in the case. Dean stated that this was the first time that he cried and told anyone about the
events that led him to prison since he was incarcerated. He thanked the therapist and then left.
The following sessions, Dean was asked to reflect on the drawn symbol, create a
sculpture out of Sculpy from it, and incorporate it with the other’s images in a group mural. He
rendered a small sculpture of a lion (Figure 7), but voiced inability to put it in an environment.
Insert Figure 7 about here
He did agree to put it on the mural paper in the corner, with its back to the drawn images. The
other members, who up until this time were quite wary of Dean, worked with him and managed
to draw radiating lines from a cross that was in the drawing, extending the lines so they fell
beneath the lion. He worked well with the group, and was able to discuss with them the theme of
the mural.
During the remaining sessions, his artwork displayed a more loose and colorful style,
culminating in the mandala he was asked to create the last session (Figure 8).
Follow-up art therapy study 14
Insert Figure 8 about here
As noted, the composition fits the entire page, and it seems to be seen from his perspective.
Unlike the previous landscape, which shows the running fence and barn in the distance, this
image portrays a sense of belonging in the environment. His BDI-II scores indicated some
decrease in depression. As well, Dean was much more interactive with the others in the group,
maintained a much brighter affect, and voiced more hope for the future, such as ideas about
working on housing projects for newly released inmates upon his parole. He projected an
extended sense of self, more awareness of his environment, and overall maintained much
brighter spirits. Thus, despite very little change in his FEATS drawing, Dean seemed to benefit
greatly from the art therapy services provided over the eight weeks.
Evaluation of the study
This project differed greatly from the pilot project the year before. Although data was
still obtained from pre- and post-session drawings, using the Formal Elements Art Therapy
Scale, the members who participated in the art therapy services were also administered the Beck
Depression Scale before and after the treatment period. This was to ascertain if there were any
changes in mood, evidence of which emerged in the pilot.
Although the data received from the pilot indicated there were significant changes, it is
believed that the results would have been stronger had an experimental group been compared to
a control group. Thus, a control group was established for this study. Inmates, who were not
scheduled to receive art therapy services during the summer, were given all pre and post art tasks
and questionnaires, as well as the assessments by the on-site senior psychologist. There were
eight weeks between the administration of the pre and the post-tests.
Follow-up art therapy study 15
However, there were some obvious shortcomings, which could have been avoided. There
were fewer participants in the experimental group than there were in the previous study, and few
participated in the control group. Future studies need to account for familiarity with the art
assessment procedures, and a longer time-period for art therapy sessions may need to be
established.
However, based on the changes perceived in the participants, and the positive effects art
therapy seemed to have on this population, this project has been greatly expanded. Additional
funding is currently being secured to expand the research to include several different prison sites,
providing the opportunity to administer the assessment tools and art therapy sessions to inmates
who have never been exposed to these procedures. As well, because of the expanded venue, it is
anticipated that there will be many more research participants. Future studies will continue to
build upon the positive findings of both the pilot and follow-up studies.
Conclusion
Upon review of the hypothesis on whether or not art therapy was effective in improving
mood in prison inmates, the answer, based on the outcome of this study, is partially yes for
depression. The results of the BDI-II and the anecdotal, qualitative data support that art therapy
was beneficial for the participants. Feedback from the correctional officers, staff psychologists
and the participants support this conclusion. However, as noted, for various possible reasons, the
results from the FEATS did not support this claim. Despite these difficulties, overall this study
seemed to provide useful evidence for the benefits of art therapy in prison. Establishing better
evaluation tools, eliminating repeat study participants to decrease familiarity with any art-based
assessment, and increasing the number of participants may strengthen the results in future
studies.
Follow-up art therapy study 16
References
American Psychiatric Association (1987). Diagnostic and statistical manual of mental disorders,
3rd. ed, revised. Washington D.C.: Author.
Baillargeon, J., Black, S.A., Contreras, S., Grady, J., & Pulvino, J. (2002). Anti-depressant
prescribing patterns among prison inmates with depressive disorders. National Criminal
Justice Reference Service. Washington, D.C.: U.S. Department of Justice.
Beck, A.T., Rial, W. Y., Rickets, K. (1974). Short form of Depression Inventory: Crossvalidation. Psychological-Reports 34 (3), 1184-1186.
Beck, A.T., & Steer, R.A. (1993). Beck Depression Inventory manual. New York: Harcourt
Brace.
Boothby, J.L. & Durham, T.W. (1999). Screening for depression in prisoners using the Beck
Depression Inventory. Criminal Justice and Behavior, 26 (1), 107-124.
Chiles, J.A., Von Cleve, E. Jemelka, R.P., & Trupin, E.W. (1990). Substance abuse and
psychiatric disorders in prison inmates. Hospital and Community Psychiatry, 41, 11321134.
Gantt, L. & Tabone, C. (1998). The formal elements art therapy scale: The rating manual.
Morgantown, WV: Gargoyle Press.
Gussak, D. (2004). A pilot research study on the efficacy of art therapy with prison inmates.
The Arts In Psychotherapy, 31(4), 245-259.
Gussak, D. (1997a). Breaking through barriers: Art therapy in prisons. In D. Gussak and E.
Virshup (Eds.), Drawing Time: Art Therapy In Prisons And Other Correctional Settings
(pp.1-11) Chicago, Il.: Magnolia Street Publishers.
Follow-up art therapy study 17
Toch, H. (1992). Mosaic of despair: Human breakdowns in prison.. Washington, D.C. :
American Psychological Association.
Winfree, L.T., Jr., Mays, G.L., Crowley, J.E., & Peat, B.J. (1994). Drug history and
prisonization: Toward understanding variations in inmate institutional adaptations.
International Journal of Offenders Therapy nd Comparative Criminology, 38, 281-295.
Follow-up art therapy study 18
Author Acknowledgements
Special thanks to Lariza Fenner, MA, Research Assistant who helped organize and coordinate
the art therapy sessions and collected the data for this study, and to Kenneth Weaver, PhD, Chair
of the Emporia State University, Psychology and Special Education Department, for his
assistance in calculating the statistics for this study.