Single-subject multiple probe designs were employed in two studies with 5 young adults who had a dual diagnosis of mental retardation and mental illness. Our aim was to determine effects of instruction designed to teach visual arts activity skills and promote personal expressiveness on acquisition, maintenance, and generalization of these skills and behaviors associated with these persons' mental health. In Study 1, a 5-second constant time delay procedure was used to teach three chosen art activities. In Study 2, an instructional package was used to promote personally expressive behaviors. After learning the skills in Study 1, participants in Study 2 displayed improvement in occurrence of behaviors associated with mental illness and increases in personally expressive behaviors.
Editor in charge: Kevin Walsh
Individuals with mental retardation have a higher incidence of mental illness than do individuals in the general population, with most investigators reporting a prevalence rate of 20% to 40% (Crews, Bonaventura, & Rowe, 1994; Thomas, 1994) as compared to 16% to 20% in the general population (Johnston & Whitman, 1990). Recognition that individuals with mental retardation can have an additional diagnosis of mental illness, also known as dual diagnosis, has developed slowly compared to many other advances in understanding mental retardation (Johnston & Whitman, 1990).
According to Crews and colleagues (1994), individuals with a dual diagnosis may be one of the most underserved groups in the United States. Galligan (1990) indicated that there is the need to develop appropriate service models for individuals with dual diagnoses (see also Matson & Coe, 1992). Thomas (1994) noted that drug therapy has been the major, and often only, approach to treatment of these individuals and that such treatment is often not appropriate. He recommended a combination of various models in treatment, including behavior therapy and alternative therapies, such as creative arts and recreation therapies.
A few programs in which creative arts therapies are utilized have been developed for individuals with a dual diagnosis (e.g., Harlan, 1991; Harlan & Hawkins, 1992; Lowenfeld, 1987; Stamaatelos & Mott, 1983, 1985; Tucker, 1993). Several case studies have documented the success of creative arts therapies in improving participants' mental health (e.g., Davis & Anderson, 1997; Edelson, 1991; Harlan & Hawkins, 1992; Katz, 1997; Lowenfeld, 1987; Stamaatelos & Mott, 1983, 1985). Descriptions of facilitated creative performances provided in these case studies are closely associated with the theory of personal expressiveness developed by Waterman (1990).
Waterman (1990) posited that individuals are selective in their pursuit of interesting stimuli, and, rather than consciously seeking emotional experience, they are more focused on pursuing specific activities. According to Waterman, individuals select activities that are personally expressive, thus enabling them to develop their potentials, capacities, and talents.
Development of Waterman's (1990) theory of personal expressiveness began with examinations of the internal consequences that make something intrinsically motivating. Intrinsic motivation has been identified as a primary dimension of leisure (Neulinger, 1981), and researchers studying intrinsic motivation have indicated that it enhances perceptions of leisure (Iso-Ahola, 1979) and is positively correlated with happiness and sense of well-being (Graef, Csikszentmihalyi, & Gianinno, 1983). Although the focus of facilitating intrinsic motivation in individuals with mental retardation has been on self-determination and choice (Dattilo & Schleien, 1994; Mahon, 1994; Mahon & Bullock, 1992; Wall & Dattilo, 1995; Wehmeyer, 1996), other constructs related to intrinsic motivation, such as personal expressiveness, warrant attention.
Individuals with mental retardation traditionally have not only been denied opportunities to make choices (Dattilo & Schleien, 1994; Wall & Dattilo, 1995), they have been denied full understanding of their potentials in terms of personal expressiveness (Lakin, 1997; Stamaatelos & Mott, 1985). Such lack of opportunities may be attributed to reasons described by Waterman (1990), namely, that (a) environments are limiting so that individuals cannot extend themselves beyond satisfying needs, (b) external social pressures influence individuals, (c) individuals become distracted, and (d) tasks are perceived as too difficult. According to Crews et al. (1994), problems contributing to mental illness for individuals with mental retardation include depression caused by lowered self-concept and stress from having to adjust to difficult social situations without adequate coping skills.
One widely recognized means of personal expression has been through the arts (Lakin, Miller, & Burkholder, 1997; Wilson, 1986). Because in case study reports investigators have specified the contribution of visual art activities to improvements in mental health (Edelson, 1991; Harlan, 1991; Lowenfeld, 1987; Phillips, 1986; Stamaatelos & Mott, 1983), it is reasonable to assume that leisure skill acquisition in art activities might prove valuable to individuals with a dual diagnosis. In addition, facilitation of personal expression, by implementing techniques outlined in arts programs (Harlan, 1991; Lowenfeld, 1987; Stamaatelos & Mott, 1983, 1985) might contribute to improvements in mental health.
In this paper we describe two studies in which we examined effects of visual arts instruction and techniques designed to promote personal expressiveness on acquisition of skills and the mental health of 5 young adults with a dual diagnosis of mental retardation and mental illness. In Study 1 we measured acquisition, maintenance, and generalization of art activity skills taught, and in Study 2 we measured behaviors associated with mental illness when techniques used to facilitate personal expression were used. For Study 1 we employed systematic instructional techniques to enable participants to acquire skills needed to progress to Study 2. Participants, setting, and materials and equipment were the same for both studies, with the exception of one less participant in Study 2.
Systematic instructional strategies have been used successfully to teach children and adults with a wide range of disabilities, including those with mental retardation, severe and multiple disabilities, and autism (Wolery, Ault, & Doyle, 1992). These strategies provide instructors with the tools to identify behaviors to be taught; to plan, implement, and evaluate instructional activities; and to determine whether desired changes occurred. Constant time delay procedure has proven to be most efficient in the use of instructional time across a range of tasks, ages, and ability levels (Wolery et al., 1992). The procedure involves the presentation of a task followed by a fixed amount of time before the instructor presents a controlling prompt. During initial trials, 0-second delay is used, with the prompt given immediately after the task direction. Then the time between the task direction and prompt is increased to a predetermined number of seconds, and constant time delay trials begin. Correct responses are those that students give consistently and correctly before the prompt. With the use of this procedure, students learn quickly, with a general error rate of less than 10% (Handen & Zane, 1987).
In this study we asked the question, What is the effect of instruction designed to teach visual arts activity skills on art activity skill (a) development, (b) maintenance, and (c) generalization by young adults (ages 17 to 21 years) with a dual diagnosis of mental retardation and mental illness?
Participants, Setting, and Materials
Five young adults with a dual diagnosis of mild to moderate mental retardation and a mental disorder, according to the definition provided by the Diagnostic and Statistical Manual of Mental Disorders (4th ed.)—DSM-IV (American Psychiatric Association, 1994), participated in Study 1 (see Table 1). They ranged in adaptive behavior age, according to the Vineland Adaptive Behavior Scale (Doll, 1965), from 6.58 to 11.33 years. Each individual was identified as having chronic difficulties related to depression or anxiety, which are two of the most frequent presenting problems of mental illness according to Burns (1989) and Young, Beck, and Weinberger (1993). Participants attended a private school for individuals with mental retardation who require extensive related services, especially psychological services. Each had an Individual Education Plan with a targeted goal of improving leisure skills. No participant had experience with constant time delay teaching procedures, and none had been given formal training in the use of art materials. A pseudonym given to each participant was used in all data-collection procedures and written documents. Informed consent and parental/guardian permission was obtained for all participants.
Instruction was given in two dyads and a one-on-one arrangement in a designated art room of the school. This room was separate from the regular classrooms and was used solely for the purposes of the research. There were no teachers or teachers' aides present during the study, and the researcher served as instructor. The room measured approximately 5.5 m × 3 m and contained a large rectangular worktable that comfortably accommodated 8 adults. Boxes of supplies were kept in the adjacent office and were placed on a corner table prior to each session. Generalization sessions occurred in participants' classrooms. Each classroom consisted of around 8 students, a teacher, and a teacher's aide and contained a desk for each student, a teacher's desk, and a rectangular table that seated 6 to 8 individuals.
Art activity supplies were kept together in a box labeled by participant. There were five activities from which participants chose the following three to be taught: mixed media drawing, watercolor, and acrylic painting. For the mixed media drawing activity, supplies for each participant were (a) three different grades of drawing pencils (hard, medium, and soft), (b) a gum eraser, (c) two black conte crayons, (d) a set of eight “classic” markers, (e) a set of eight soft colored pencils, and (f) pad(s) of at least 10 sheets of 30.5 cm × 45.7 cm drawing paper. For the watercolor activity the supplies were (a) a set of 10 watercolors in plastic tubes, (b) 4 sable brushes of different sizes, (c) a plastic watercolor palette, (d) pad(s) of ten 30.5 cm × 45.7 cm watercolor paper, (e) a 338-liter plastic cup, and (f) paper towels. For the acrylic painting activity the supplies were: (a) one No. 2 pencil, (b) eight tubes of different colored acrylic paints, (c) four acrylic paint brushes of different sizes, (d) one plexiglass 20 cm × 25 cm palette, (e) 10 or more 30.5 cm × 45.7 cm canvas boards, (f) an empty 608-liter coffee tin, and (g) paper towels. The instructor prepared data-recording materials prior to each session. In addition a video camera was used to record sessions.
A 5-second constant time delay procedure was used to teach art activities to participants, 4 of whom were in participant dyads. The dyad arrangement facilitated individualized instruction that cannot be addressed in a larger group yet enabled the participants to gain experiences in working with others during art activities. Because there was an uneven number of participants, the remaining individual was taught one-on-one. Daily sessions (30 minutes) consisted of two trials of the total chained task for the activity being taught. Each chained task was divided into three parts: (a) set up, (b) use of materials, (c) clean up. One participant in the dyad was taught set up and clean up while the second participant observed. The second participant was taught to use materials while the first participant observed. Observation and direct instructional tasks were counterbalanced between trials so that each participant received direct instruction on each step of the task across two trials in a session. In the one-on-one arrangement, the total chained task was taught in two trials at each daily session for all three art activities. Criterion for each participant for each chained task was 100% unprompted correct responses when reinforced on a VR3 (variable ratio reinforcement schedule, occurring every third step) across two consecutive trials in a dyad, and an FRx (fixed ratio reinforcement schedule) at the last step of each task analysis across two trials in the one-on-one arrangement. Once a dyad had reached criterion for an art activity and baseline for the next activity had stabilized, the next activity in the series was taught.
Experimental Design and Procedures
A single-subject multiple-probe design across tasks was replicated across 5 participants (Tawney & Gast, 1984). In multiple-probe single-subject design, direct and systematic replication of identical conditions over time enhances internal validity, with participants serving as their own controls (Dattilo, Gast, Loy, & Malley, 2000). Thus, behavior changes occurring with introduction of an intervention can be attributed to the intervention rather than to extraneous variables (Kazdin, 1982).
The instructional program consisted of (a) baseline probe, (b) instructional condition using constant time delay procedures, (c) maintenance, and (d) generalization. Baseline probe sessions occurred prior to instruction for each art activity. After a dyad reached stability by maintaining the same number of correct responses across three data points for one art activity, instruction using 5-second constant time delay began for that activity. Probe sessions continued until stability for the other two activities occurred. Upon reaching criterion in the first instructed art activity, participants in the dyad received probes for all three activities. They then received instruction using 5-second constant time delay in the second art activity, with periodic maintenance probes occurring on the first activity. Upon reaching criterion in the second art activity, they were given probes for all three activities. Following these probes they received instruction using 5-second constant time delay in the third art activity, with periodic maintenance probes occurring on the first and second activity. Upon reaching criterion in the third art activity, participants received probes for all three activities. Then, in a final condition we evaluated generalization across settings and instructors.
Experimental control was demonstrated with systematic introduction of the 5-second constant time delay instructional procedure to each participant. Effective control was indicated when participants' unprompted correct responding on targeted steps increased following constant time delay instruction and did not increase during preinstruction probes.
The first two trials of the instructional condition were conducted using a 0-second time delay and continued until both participants of the dyad reached 100% correct prompted responses across two trials. Each trial consisted of all of the steps in the task analysis. Subsequent trials were conducted using a 5-second time delay. The instructor's controlling prompt for each step of the task analysis was a model coupled with verbal description of the response. Reinforcement for correct responding consisted of descriptive verbal praise on a continuous reinforcement schedule (CRF after each correct response) for the 0-second time delay trials and for the 5-second time delay trials until 100% correct responding was reached.
For the 0-second delay trials, the instructor indicated for which tasks each member of the dyad would be responsible and delivered an attentional cue, “Mr.______ and Ms.______, are you ready?” After hearing affirmative responses, the instructor provided the prompt for the first step of the task analysis to the participant receiving the instruction. Upon correct responding by the participant receiving instruction and attending by the other participant, the instructor provided verbal praise to both. The controlling prompt for the next step was provided immediately after the reinforcement, and each step in the task analysis was taught as in the first step until completion of the art activity. In two trials of the entire chain of steps, each participant performed all of the steps in the task analysis in the dyad arrangement. After both participants reached 100% correct prompted responses across two trials during the 0-second time delay, trials using 5-second time delay began. When participants indicated that they knew their roles and were ready, following the attentional cue, the instructor provided a general task direction, such as, “Paint a picture.” The instructor then waited 5 seconds to enable the participant to initiate the activity before the controlling prompt. After providing the appropriate consequence for the first step, the instructor waited 5 seconds for initiation of the second step and provided a consequence based on the response. The consequence for correct responses was “Keep going,” if necessary, and for incorrect or no response, the instructor modeled and verbally described the correct response. Instruction continued for all subsequent steps in the task analysis as in the first two steps. Participants were not prompted when it was time to switch roles, and prompts to observe were not provided to the observing participant.
In addition to the instructional program, baseline probes for Study 2 occurred in Study 1. Behaviors associated with mental illness, product evaluations, and participant product descriptions (see Study 2 for description) were measured once in each probe condition prior to each art activity instruction and in the final probe condition.
Maintenance and Generalization
Following instruction in each of the art activities, participants received one-on-one maintenance sessions once a week or each art activity taught until the remaining art activities were taught. The procedures for the maintenance sessions were identical to those used during 5-second time delay instruction with a fixed ratio reinforcement schedule at the last step of each task analysis. If less than 100% unprompted correct was recorded, the instructor was to use a variable ratio reinforcement schedule (VR3) at the next trial for that task analysis and continue until 100% unprompted correct was achieved. The instructor would then return to a fixed ratio reinforcement schedule (FRx) at the last step of each task analysis. At the completion of the instructional condition for the last art activity learned, final probes were taken on all three art activities. Then Study 2 began, which enabled further maintenance for the activities, with maintenance probes occurring once a week.
To assess generalization across people and settings, we asked participants to take their supply boxes to their classrooms once a week for 2 weeks once the instruction phase had ended for a given art activity. A teacher who had been trained in the 5-second time delay procedures asked the participant to demonstrate the steps in completing the art activity. The teacher used a fixed ratio reinforcement schedule at the last step of each task analysis, with the number of unprompted corrects recorded.
A continuous event-based recording system (a symbol was recorded for each response for each step of the task analysis) through direct observation was used to record each participant's motor responses during art activities (Wolery, Ault, & Doyle, 1992). Correct responses were recorded as either unprompted corrects or prompted corrects. Incorrect responses were recorded as unprompted (or nonwait) errors, prompted (or wait) errors, and no response. Performance on each step was recorded one time in each session at the beginning of the session. Participant performance was monitored throughout each condition.
Accuracy and Reliability
To establish a standard for coding procedures, the instructor reviewed five 2- to 5-minute videotaped sessions of participants being taught a simple motor activity using 5-second constant time delay several times. Using these videotapes, we trained an observer to follow coding procedures until 90% agreement with the established standard was achieved on three consecutive occasions. The observer recorded data from the videotapes every fifth session to monitor reliability of instructor observations. A point-by-point method was used to calculate interobserver agreement percentages. This method is the number of participant response agreements divided by the number of response agreements plus disagreements and multiplied by 100. As suggested by Tawney and Gast (1984), at least 85% interobserver agreement was required to continue. For procedural reliability, the instructor's accuracy in administering the instructional program was assessed. Data were collected at least one time a week or at least one time in each condition. Each session to be assessed was videotaped and an observer used the videotape to measure instructor behaviors. In addition to the recorded data, the observer provided detailed verbal feedback to the instructor after each observation.
Visual analysis of graphic data was used to determine results (Tawney & Gast, 1984). Each participant independently and with 100% accuracy completed the three art activities on two consecutive occasions within 30 minutes (see Figures 1 through 5). The mean number of sessions to VR3 criterion was five for drawing, five for watercolor, and four for acrylic. There was a mean of four errors to criterion for drawing, four for watercolor, and two for acrylic. The mean number of minutes per session to VR3 criterion was 30 for drawing, 38 for watercolor, and 29 for acrylic. Absolute level change refers to the absolute difference in point values from the last plotted point in one condition to the first plotted point in the next condition (Tawney & Gast, 1984). Absolute level changes from baseline to intervention ranged from 10% to 75%, with changes decreasing with each successive instructional activity.
Reliability of the observations by the instructor was within a range of 91% to 100% agreement, as monitored by an outside observer. Procedural reliability, determined by an observer assessing the instructor's accuracy in administering the instructional program, ranged from 86% to 100%.
All three research questions were answered; however, effects of the instructional program on skill acquisition could not be determined because baseline data were accelerating for many skills. The art activities (tasks) selected by participants were not totally independent; that is, many of the steps were repeated in each task analysis. Therefore, effective control could not be indicated. Participants' unprompted correct responding on targeted steps increased following constant time delay instruction, but it was increasing during preinstruction probe sessions in most cases. However, each participant learned three art activity skills and maintained and generalized these skills, steps necessary to begin Study 2.
Our goal in Study 2 was to determine the answer to the question, What is the effect of the instruction designed to promote personal expressiveness on the mental health of young adults with a dual diagnosis of mental retardation and mental illness?
Participants, Setting, and Materials
Four adults from Study 1 participated in Study 2. Ms. Mann was absent from the school for many days at a time during the second study and withdrew from the study. Instruction occurred in a group with all participants in the same art room containing the same materials and equipment used in Study 1. In addition, participants could choose from a variety of papers and collage materials located in the room. No other activities were available for the participants.
An instructional package consisting of seven techniques was used to promote personally expressive behaviors. Group sessions of 45 to 60 minutes were held daily for approximately 4 weeks. Participants chose materials they wanted to use from the three art activities learned as well as from additional materials located in the art room. Each session consisted of using art materials in appropriate ways, based on skills learned in Study 1, with expression facilitation techniques employed by the instructor.
Expression facilitation techniques
Procedures based on the approaches of Lowenfeld (1987) and Harlan (1991) were designed to facilitate personally expressive behaviors. Seven techniques were used. Stimulations are verbal suggestions to focus participants' attention on their feelings about their bodies that were intended to enhance their awareness of how they communicated. Any participant movement could be used. For example, if a participant yawned, a statement was made that she yawned and then questions about how it feels to yawn were asked: “Are you tired? How does your face feel when you yawn?” Kinesthetic experiences are suggestions or demonstrations for kinesthetic experiences that were directly applied to the use of art media. For example, participants were asked to move their arms up and down, as demonstrated, and then asked to hold the pencil on a large sheet of paper and repeat the same movement. Use of feelings involved recognizing participant feelings and encouraging them to use their feelings in the art activities. For example, if participants expressed anger about something that recently happened, they were asked to show in the art what the anger looks like. Representation of things (abstract designs) was encouraged if attempts at realism seemed frustrating. For example, if participants expressed that they wanted to draw a horse but stated “I can't draw a horse,” then suggestions for thinking about the horse were offered: “Think about what a horse looks like when it is running. What sounds does a horse make? How can you show that using lines and shapes?” Elaboration of designs was encouraged. For example, if participants quickly drew repetitive designs, suggestions for elaborating on the designs might have included coloring in all of the white spaces, or adding another type of pattern on top of the first one. Content-based comments were used. Rather than repeatedly praising the art, the instructor directed comments toward the actual content and method used in the work, which reflected a genuine interest in the work and the process involved. For example, typical comments might be, “Can you tell me about that cow you just drew?” or “I see that you are painting with watercolors today. They seem to work well for you when you paint landscapes.” Nonjudgmental acceptance involved reassurance of the art produced by participants. Silence was used as a way of communicating acceptance; it was not necessary to provide positive comments on every aspect of a work in progress. Reassurance was provided when participants expressed difficulty, by asking which parts of the art participants liked and why and which parts they disliked and why.
These seven techniques were used systematically throughout the intervention condition in a sequence suggested by Lowenfeld (1987). Each of the first six techniques was used at least once with each participant during designated sessions as follows: stimulations and kinesthetic experiences during Sessions 1, 2, and 3; use of feelings during Sessions 4, 5, and 6; representation of things during Sessions 7, 8, and 9; elaboration of designs, during Sessions 10, 11, and 12; and content-based comments during Sessions 13 through 15. Nonjudgmental acceptance was used at least once with each participant during each session.
Behaviors associated with mental illness
Six behaviors were measured, based on descriptions in the DSM-IV. According to this manual individuals with chronic mental illness exhibit a high incidence of depression and general anxiety disorder manifested by (a) negative affect or frowning, (b) disengagement from an activity, (c) negative references to self, (d) negative verbalizations towards others, (e) verbalizations related to death, and (f) psychomotor agitation. Negative affect was observed in facial expressions when the inner corners of the brows were drawn obliquely upward and together, the eyes were narrowed, and the corners of the mouth were pulled downward (Izard, 1991). Disengagement was identified when participants did not manipulate any materials associated with the art activity and were not physically attentive to the art activity during the 5-minute interval. Physical attention referred to facing and looking at the art but did not require active manipulation of materials; however, if participants remained totally still, not moving for the entire 5-minute interval, disengagement was identified as well. Negative references to self were observed when participants made statements that were negative in relation to their abilities or appearance, such as “I'm not good at drawing,” and “I have ugly hair.” Negative verbalizations towards or about others occurred when participants made negative statements directed to others in the group or negative references to others in general, such as “I don't like you,” ” You are really ugly,” “People are always mean to me,” and “She keeps bothering me.” Verbalizations related to death were statements that indicated that they would like to die. If this behavior were to occur during any 5-minute interval, it would be reported immediately to the staff psychologist or psychiatrist. Psychomotor agitation occurred when participants exhibited motor movements of any part of the body that were not needed in the art activity and that were repeated at least once every 10 seconds for six consecutive 10-second intervals (or one minute). Examples of movements included shaking head; flicking hands; tapping fingers; tapping a pencil or other instrument; shaking a pencil or other instrument; shaking a foot, leg, or another body part; wringing hands; pulling, scratching, or rubbing skin, clothing, or other objects; rocking in a chair; getting up and down out of a chair; pacing back and forth; and rocking back and forth while standing.
Other measures were used to determine whether the interventions were effective tools for facilitating personal expressiveness. These measures were (a) product descriptions, which evaluated the richness or quality of the art work; (b) personal product description, which used participants' descriptions; and (c) the amount and kind of self-selection occurring outside of the sessions.
Experimental Design and Procedure
A single-subject multiple-probe design was used across 4 participants, consisting of (a) baseline and (b) Intervention 2 (goal was to promote personal expressiveness). Behavioral observation was the primary method of data collection, with interviews and evaluations of visual art used to determine personal expressiveness. Additional interviews and questionnaires were used to assess the intervention's social validity. Baseline probes of the behaviors associated with mental illness occurred prior to facilitation of expressive behaviors. After behavior stability was reached, introduction of the intervention was staggered across 1 or 2 participants at a time. Each participant engaged in the intervention for at least 15 sessions.
In addition to measures of behaviors associated with mental illness, product evaluations and participant product descriptions were conducted at least once during baseline and once a week during intervention as indicators of personal expressiveness. The instructor kept a log to record the self-selection that occurred outside of the sessions.
Experimental control was demonstrated with systematic introduction of personal expression facilitation techniques to each of 2 participants and 1 dyad. Effective control was indicated when participants' targeted mental health behaviors decreased and personal expressiveness indicators increased during intervention and not during preintervention probe sessions.
Participants' behaviors reflective of mental illness and indicators of personal expressiveness were assessed prior to introduction of the expression-facilitation techniques. In Study 1 behaviors were measured, and descriptive product evaluations and participant product descriptions were obtained once during each probe condition prior to instruction in the art activities and at the completion of instruction. In Study 2 at least three probe sessions of behaviors associated with mental illness with each participant occurred in the initial probe condition until stability was reached. Daily sessions in the initial probe condition consisted of using the supplies designated for the three art activities for approximately 20 minutes, without intervention from the instructor. Once stability in behaviors associated with mental illness was reached for 1 participant, the intervention was introduced to that participant. Probes continued with the other 3 participants until stability was reached. Once stability was reached for 2 more participants and the first participant had attended at least three sessions, the second 2 participants began the intervention. Once stability was reached for the second 2 participants and the first participant had attended at least six sessions, the last participant began the intervention. Thus, introduction of the intervention was staggered across 4 participants. Probes occurred once every three intervention sessions for participants who met the baseline criterion but had not yet begun the intervention.
Each session in each condition was videotaped. Occurrence of behaviors associated with mental illness was recorded through observations of videotapes by an observer using a 5-minute partial-interval time-sample recording system. Using this system, we divided each of the 15 sessions on videotape into equal intervals, and occurrence or nonoccurrence of the behaviors was recorded once for each interval (Tawney & Gast, 1984). Occurrence of the behaviors was assigned a value of one and nonoccurrence, a value of zero. For psychomotor agitation, six consecutive occurrences in 1 minute were needed to count as one occurrence in a 5-minute interval. Total points for each session were plotted on graphs.
Indicators of personal expressiveness were recorded by an observer using a checklist to measure descriptive product evaluations, and participant product descriptions were obtained through open-ended interviews. For each of these two measures, the product was completed artwork produced during that week and chosen by the participant. The product evaluation checklist was developed based on elements of art structure as interpreted by Ocvirk, Bone, Stinson, and Wigg (1968). The checklist evaluated variety of line, shape, value, and texture, use of color, and overall quality and organization of the product. Items were tallied and a percentage calculated. Interview guidelines for participant product descriptions were based on guidelines for eliciting personal reflection of artwork provided by Rubin (1984). Five open-ended interview questions were asked: (a) Tell me about your picture. (b) Does it have a story to go with it? (c) What were you thinking about when you were doing this? (d) As we look at it now, what does it remind you of? (e) Do you want to add to it or change anything? If so, what? The number of times self-references were made was tallied for each interview and compared from one session to the next. Descriptive product evaluations and participant product descriptions enabled evaluation of personal expressiveness over time. Each measure was recorded once a week throughout each condition.
Because individuals are intrinsically motivated to select activities that are personally expressive (Waterman, 1993), we used self-selection of art activities as an additional indicator of personal expressiveness. The instructor recorded participant requests to engage in the activities outside of the regular sessions, requests for supplies to take home and use, reports of art activities initiated in the classroom and at home, and actual art products completed by participants and brought to show the instructor. A continuation of the maintenance procedures for the Study 1 instructional program was followed, with maintenance probes occurring once a week throughout each condition.
Accuracy and Reliability
A 15-minute videotaped session of participants in a group art activity was reviewed by the instructor to establish a standard for coding behaviors associated with mental illness. Next, an observer who was not involved with other aspects of the study was trained and repeatedly viewed the videotapes and followed the coding procedures until 90% agreement with the established standard was achieved on three consecutive occasions. For product evaluation accuracy, the instructor evaluated three visual art products using the product evaluation checklist to establish a coding standard. Next, an artist was trained to use the checklist and evaluated three products repeatedly until 90% agreement with the established standard was achieved on three consecutive occasions.
For the behaviors associated with mental illness, a primary observer recorded data from each videotaped session. A second trained observer viewed videotapes of every third session and recorded data on behaviors associated with mental illness. For reliability of product evaluation, an artist used the product evaluation checklist to evaluate two products by each of the 5 participants A point-by-point method was used to calculate interobserver agreement. If less than 85% agreement occurred, the measurement procedures were examined to determine the discrepancy and correct the procedures.
Instructor's accuracy of facilitation techniques implementation was assessed at least once a week, one time in each condition, or one time for each instructor technique by using videotapes. The observer was trained by watching videotapes of the instructor using the facilitation techniques with a participant. Accuracy of the use of the seven facilitation techniques with each participant was recorded. Procedural reliability was estimated by dividing the number of uses of the designated facilitation techniques by total number of uses, missed opportunities, and inaccurate uses multiplied by 100.
Mental Health Indicators
There was consistent improvement from baseline in occurrence of behaviors associated with mental illness for all participants, with a decreased percentage of behaviors during the intervention. During baseline measures each participant exhibited one or two indicators in each session, with an overall mean of 12%. Although baseline percentages were low, each participant maintained an average of only 1% during intervention. Mr. Downing displayed four behavioral indicators during baseline, most notably negative verbalizations, which diminished during intervention. Ms. Allen consistently exhibited negative affect during baseline, which greatly diminished during intervention. Mr. Randolph showed negative affect and occasional negative verbalizations, which were not as apparent during intervention. Mr. Chambers exhibited negative affect, negative verbalizations, and psychomotor agitation during baseline, and only negative verbalizations during intervention. At most, only two of the six behaviors were exhibited in a 5-minute interval, but usually only one behavior, hence the low baseline scores. Examples of negative verbalizations are: “Some people gonna get stung by a scorpion,” and “Hey,______ [name], ______ [name] said a P word behind your back yesterday. … I better be leaving this place so you won't call me that P word no more.”
Changes from baseline to intervention for product evaluations were not consistent across participants. Ms. Allen had the highest absolute level change, with 16%. Mr. Chambers had an absolute level change of 11%, which was a reduction from baseline, but he had an increase of 25% when comparing baseline and intervention mean scores. Mr. Downing had only a 5% absolute level change from baseline to intervention, and his mean baseline intervention scores remained the same at 58%.
Participant Product Descriptions
Participant descriptions of products revealed the level of personal reflection and investment. Participant statements related to themselves when describing products were tallied. A comparison of number of self-references made prior to intervention and during intervention was made. (See Table 2 for number of self-references in each product description.) Mean number of self-references in product descriptions from baseline to intervention increased for all participants except Mr. Chambers, who showed a slight decrease in number of self-references
An example of Mr. Downing's descriptions during baseline was: “It's black and white stripes. Well, they are the stripes on a zebra. The zebra is a wild animal of Africa.” During the intervention, he gave the following description:
I might call this a reality but I just made it. I like this the best of the two pictures I did, because of the old-fashioned colors. It's like a rainbow triangle. It makes me feel good, I'm pretty happy.
During Ms. Allen's baseline, she provided the following: “I don't know what it is. The leaves, when it's fall, they change colors.” One of her descriptions during the intervention was as follows:
I like it. The reason why I got bright colors is because it is summer. The picture makes me happy. I'm happy about summer and graduating, that's why I made this. It's just about feelings. It reminds me of my room.
All participants displayed self-selection behaviors early in the studies. Mr. Randolph indicated that he wanted to bring his supplies home in the fifth session of Study 1. When he completed drawing instruction, he was carrying his drawing supplies with him everywhere and had begun to produce art outside of sessions. Once he completed watercolor instruction, he asked the instructor to go to the art supply store so that he could purchase watercolor paper and replenish drawing supplies. With the instructor's assistance, he made a list of items he wanted and estimated cost, supplies were purchased, and he continued to produce art outside of sessions on almost a daily basis according to his and his teacher's reports. Ms. Allen requested to bring drawing supplies home in the fifth session of Study 1. During instruction she completed approximately five drawings in her sketchpad. Once she had learned acrylics, and was being probed for 100% correct responses (Session 23), she asked when Study 2 sessions would begin so that she could have more choices in use of media. In the fifth session of the Study 2 intervention, she asked to bring watercolor supplies to her classroom and started a watercolor that she finished in the next session. After that session she began to bring products back and forth between sessions and the classroom or leave them in the art room and come at an unscheduled time to work on them. Unlike Mr. Randolph and Ms. Allen, Mr. Downing did not request to bring supplies home. However, once he had mastered the drawing skills, by Session 6, he wanted to continue to produce art beyond the length of the session. This behavior continued throughout Studies 1 and 2. He would complete a product, then request to start another, and yet another. He was allowed to stay in the art room after many of the sessions to complete paintings during both Studies 1 and 2. Although Mr. Chambers did not request to bring supplies home or to his classroom, he brought colored pencils to his classroom from home to use during the third week of Study 1 and requested use of a sketchpad. Once he had learned drawing skills he continuously wanted to produce art beyond each session in both studies. He remained in the art room after many sessions to complete paintings, and frequent reminders were needed daily to persuade him to return to his classroom or go to his part-time job. Ms. Mann withdrew from Study 2, but she demonstrated self-selection behavior in Study 1. She was given drawing supplies to take to her classroom after Session 11. Her teacher required that her supplies be kept locked in a drawer due to some problem behaviors of her classmates. She was told that the supplies would be there if she needed them, but she did not self-select the supplies during her free time, possibly because she forgot that they were available. However, about 50% of the time she wanted to stay beyond session times to finish a product. Once her products were completed, she would show them to everyone she passed on her way to her classroom, enthusiastically indicating, “Look what I did today!” She expressed a strong interest in continuing in Study 2 but was absent from school frequently due to circumstances beyond her control. She joined the group when present, but could not be a part of the intervention because of the systematic approach required for data collection.
Accuracy and Reliability
For reliability of observations of behaviors associated with mental illness, agreement between the observer and instructor was in a range of 90% to 100%. For reliability of product evaluation, an artist used the product evaluation checklist to evaluate two products by each participant. The range of agreement was 89% to 100%. To measure instructor accuracy in the use of facilitation techniques, an observer colleted data five times, with a result of 100% accurate use of the techniques.
Social validity was determined by examining (a) the social significance of the instructional program and facilitation techniques, (b) the appropriateness of the procedures, and (c) the effects of the program on participants (Wolf, 1978). Three different procedures were used to monitor social validity. Significance of instructional program and facilitation techniques were determined by a review of the planned methods by (a) an art educator who works with adults who have mental retardation, (b) the education director at the research site, and (c) members of the research team.
Appropriateness of the procedures
A social appropriateness questionnaire with five open-ended questions was administered by the instructor six times to each participant throughout Studies 1 and 2, an average of once every 6 or 7 sessions. The questions were: (a) How do you like the art instruction (activities) so far? (b) What do you like most about the art instruction (activities)? (c) What do you like least about the art instruction (activities)? (d) What do you think that we could do differently to make it better? (e) Do you have any other suggestions or comments about the program? Answers to these questions determined participant satisfaction with the procedures. Suggestions for change that were feasible and did not interfere with the research design were to be implemented. Out of a total of 140 open-ended questions, only 3 answers expressed discontent. Answers tended to be very short, for example: “I like it.” “So far, so good.” “I like this group.” “It's great. It's fun too, and it's never boring.” “All is good.” No suggestions were made about how to improve the program, even though the questions were phrased in different ways that would be understandable to participants (Dattilo, Hoge, & Malley, 1996). All respondents indicated that they liked the program the way it was.
Social importance of the effects of the program
Social importance questionnaires were administered to an individual significant to each participant, to the education director, and to each participant's teachers on the first day of Study 1 (12 completed) and the last day of Study 2 (14 completed). The questions were: (a) Do you think the individual will benefit (has benefited) from the art program? (b) Do you think the individual will learn (has learned) a skill that he or she will be able to continue to use? (c) Do you think the program will positively affect (has positively affected) the individual's mental health? (d) Do you think the program can be of value to people with disabilities? Response categories were not at all, a little, yes, and very much. On the pretest 12 individuals responded positively by indicating yes or very much to a total of 48 statements 90% of the time and responded less favorably by indicating not at all or a little 10% of the time. On the posttest, 14 individuals responded positively to a total of 56 statements 96% of the time and responded less favorably 4% of the time. Although each social importance questionnaire included a space for comments, few comments were made on initial surveys. However, on the final surveys each individual wrote positive, often lengthy, comments. For example: “The program was extremely valuable for ______.” “ She looked forward to going to class and was always proud of her work.” “ The program provided ______ with a way of expressing herself in a positive way.” “The program gave this student a sense of involvement …made him proud of something he had done.” Modified questions from the questionnaires in the form of interviews were presented to each participant. Responses were consistently positive and similar to responses to the appropriateness of procedures.
Results of Study 1 indicated that although many baseline measures showed acceleration in skill acquisition, 5 young adults with a dual diagnosis of mental retardation and mental illness acquired art activity skills using 5-second constant time delay instructional procedures. Because the skills needed to engage in the chosen art activities were not independent, skill acquisition for one activity contributed to accelerated baseline for the next activity. Possibly, constant time delay procedure facilitated a faster rate of skill acquisition and, in part, accounted for the 100% maintenance and generalization. Motivation to engage in the activities was high, as indicated by unexpected early self-selection behaviors, and might also have contributed to the consistently high maintenance and generalization scores. Plans were to record self-selection behaviors during Study 2, but because participants began requesting use of the materials outside of the sessions during Study 1, a daily log was kept to record these behaviors. This high level of self-selection indicated that exposure to the activities during probes and early skill acquisition contributed to a desire to continue the activities as a means of personal expression. According to Waterman (1990), individuals are quite selective in their pursuit of interesting stimuli, and they select activities that serve to further develop their talents and potentials.
Results of Study 2 indicated that although baseline measures for occurrence of behaviors were low, the instructional program had a positive effect on the mental health of young adults with a dual diagnosis of mental retardation and mental illness. Intervention strategies were based on Lowenfeld (1987) and Harlan's (1991) approaches to facilitate personal expressiveness. Because each technique was used effectively in case studies by either Lowenfeld or Harlan, results of this study support their work. Whether receiving attention from the instructor influenced behaviors, apart from the intervention strategies, could not be determined. By following the strategies closely, the instructor attempted to remain as neutral as possible throughout the sessions. When not implementing the strategies, the instructor remained quietly at the table. As observed on the tapes, participants seemed highly engrossed in the activities and often unaware of the instructor's presence.
Results of product evaluations were inconclusive relative to personal expressiveness. The quality or richness of the products, as measured by the product evaluation, was not influenced by the intervention. The relationship between the subjective experience of personal expression and quality of art produced was not determined by this instrument and warrants further exploration.
Evidence of an increase in personal expressiveness emerged from participant product descriptions, which involved self-references. Waterman (1990) indicated that self-actualization is the key to the selectivity involved in personal expressiveness. In her work with older adults with mental retardation, Harlan (1991) indicated that producing art enabled the inner experience to be more accessible. The self-references during the intervention by all participants were very positive and, with the exception of Mr. Chambers, increased in number from baseline to intervention. An explanation for the lack of increase in self-references by Mr. Chambers might be that he already had an interest in the art activities that might serve as a basis for personal expressiveness during baseline. He had personal supplies (colored pencils) brought from home that he kept in his classroom. Perhaps he had previously selected drawing as an activity that was personally expressive for him, and the other two art activities became a natural extension of the benefits that he derived from drawing.
Although conclusions about personal expressiveness cannot occur based on self-selection behaviors and self-references, Waterman (1990) indicated that personally expressive activities continue to be selected over time and serve as a personal narrative. According to Waterman, individuals select activities that enable them to further develop their potentials, capacities, and talents. Although individuals with mental retardation might not have the cognitive abilities to understand whether activities serve to develop their potentials, in this study participants' self-selection behaviors and increased self-references from baseline to intervention indicated a desire to continue in activities.
Increased interest and a high frequency of self-selection behaviors were reported in a case study by Stamaatelos and Mott (1983, 1985). They described a girl with severe mental retardation who participated by choice in their programs and began talking more positively about herself, dressing more neatly, and participating more in social and academic activities. Whether she possessed cognitive abilities to understand self-actualizing components of her activities is not known; however, she selected activities over time that served to develop her talents and contributed to positive mental health.
If individuals with mental retardation and mental illness are given the opportunity to develop activities that might serve to further enhance their skills and their ability to express themselves, it is possible that they may experience improved mental health. As Lakin (1997) indicated, professionals working with such individuals must break out of their “traditional professional molds to recognize that generalized low expectations for people with developmental disabilities have been their primary impediment in so many areas” (p. 5), particularly in those involving expression and the arts. In the continuum of community supports provided to people with mental retardation and mental illness, opportunities to engage in expressive activities need to be included as essential components of a quality life, as essential as housing, employment, and social activities.
Note: This study was completed in partial fulfillment of the first author's doctoral degree from the Department of Recreation and Leisure Studies at the University of Georgia. Funding was provided, in part, by the National Therapeutic Recreation Society. Appreciation is extended to the participants, facility personnel, Nancy Tankersley, Andrea Rea, Douglas Kleiber, Gail Hoge, and Andra Nyman at the University of Georgia.
Authors:Sharon M. Malley, EdD, Independent Consultant/Researcher, Within Sight, Inc. 5612 2nd St., Arlington, VA 22204 ( firstname.lastname@example.org). John Dattilo, PhD, Professor, Department of Recreation and Leisure Studies, and David Gast, PhD, Professor, Department of Special Education, University of Georgia, Athens, GA 30602