Adults with mental retardation were assessed for their ability to use visual tools for identifying body parts. Participants were shown three representations: anatomical dolls, anatomical drawings, and live models, each of which had a sticker placed on a body part. They were asked to name that body part and place a sticker in the same place on their own body. Results indicated that verbal labeling was easier for participants with mild mental retardation compared to moderate mental retardation. Level of mental retardation affected the participants' ability to correctly place the stickers. Form of representation was also important. Live models were easier to use compared to dolls and drawings. Implications for forensic evaluations are discussed.
Editor in charge: Steven J. Taylor
People with mental retardation are at an increased risk of sexual abuse when compared to the general population (Furey, Grandfield, & Karan, 1994; Sobsey & Varnhagen, 1991; Westcott, 1993). One reason this occurs is that individuals who have mental retardation are less likely to be able to report accurately about incidents they have experienced. Some researchers have advocated techniques to improve this reporting (e.g., Baladerian, 1992; Valenti-Hein & Schwartz, 1995). These techniques include the use of specific questioning methods as well as the use of visual tools. In this paper visual tools refers to the use of any tool to help a person describe what happened to them. Common visual tools used for this purpose are anatomically detailed dolls and anatomical drawings.
Anatomically detailed dolls have been the source of controversy in reporting of sexual abuse among children. A large body of literature has addressed many of the advantages and concerns (see Everson & Boat, 1994). Anatomical drawings have also been used in sexual abuse interviews and have raised some controversy as to whether they are leading because they depict unclothed individuals (Groth, 1990). Perhaps the most important concern for both of these tools is whether they actually increase the reporting accuracy of the person using them. In research with children of average intellectual ability, Saywitz, Goodman, Nicholas, and Moan (1991) suggested that use of dolls does improve reports. Similar research has been conducted using drawings with comparable results (Kelley, 1985).
A concern that exists for the use of both dolls and drawings is that individuals reporting abuse with these tools may not have adequate generalization skills to identify the correspondence between the tool and themselves. If dolls or drawings are used to report something that occurred to the person's body, then it is important that the relationship between the part of the doll or drawing indicated corresponds cognitively to the part of the body the person believes was touched.
Recent exploratory studies with children suggest that this representation skill is something that is gained over time. In a series of studies DeLoache and colleagues (DeLoache, 1995; DeLoache Miller & Rosengren, 1997; Smith & DeLoache, 1995) suggested that young children under 4 years of age may not have developed the cognitive ability to perform this task. Using a dot to sequentially mark various body parts on an anatomically detailed doll, these authors found that older children were more able to place the dot on a corresponding body part than were younger children. This was not, however, related only to a developing ability to identify body parts. Both younger and older children were more able to place a dot on their own bodies when the demonstration dot was placed on a live model. Thus, generalization from the doll to their own bodies appears to be at issue.
If the cognitive complexity of representationalism is an issue, people who have mental retardation might also find this task difficult. Dolls have not always been admitted into testimony of adults with milder forms of mental retardation. For example, in State v Scherzer, Scherzer, Grober and Archer (1992), defense attorneys argued successfully to disqualify this form of testimony by stating that because the reporter was verbal and thus able to report events, the use of dolls would be prejudicial as it might convey an inappropriate concept to the jury that the alleged victim was younger than their chronological age. In general, adults with mental retardation who are nonverbal are likely to be more severely impaired in cognitive functions.
Thirty individuals were selected from a sheltered workshop setting to participate in this study. They ranged in age from 24 to 53 years and were chosen from two levels of intellectual ability. Half were functioning in the mild range of mental retardation and half were functioning in the moderate range. Information about functioning level was obtained from standardized testing included in their charts from the workshop. All participants were their own guardians, and all were informed about the research and agreed to take part. They were paid $2.00 for their participation.
One set of Hispanic anatomically detailed dolls (Anatomical dolls, 1985) and one set of anatomical pictures (Groth, 1990) were used as the visual tools. Round blue stickers were employed to identify body parts. A standard protocol that indicated the presentation order was used to obtain responses (see Appendix A). The recording sheet is shown in Figure 1.
Participants were told about the study and signed a consent form, which was reviewed orally and all questions answered before the participant signed. They were then presented with one of three representations: live models, anatomical dolls, or anatomical pictures. A Latin-square design was used to randomize the order of presentation of the representations. With each form of visual tools presented, a series of 20 trials was used. All 20 trials for that representation were completed before the next representation was presented. Presentation order was matched for adults with mild and those with moderate mental retardation so that equal numbers from each disability level received each of these presentations.
An attempt was made to preserve ecological validity by using visual tools according to commonly accepted practices. Participants received the doll that matched their gender and age, as suggested in protocols that use anatomical dolls for child abuse assessment (MacFarlane & Krebs, 1986). The picture the participant received was chosen according to the directions provided for use of this visual tool in forensic settings (Groth, 1990). Thus, all males were given the male adult picture and doll to use for the representations. All females were given the adult female doll and picture. The live model used for all participants was female.
Participants did not watch as the sticker was moved from one body part to another. They were then asked to name the part of the body where the sticker was placed. Stickers were placed over the clothes of the anatomical dolls and the live models. Participants were then instructed to place a second sticker on their own body in the same place. Accuracy of placement was recorded, with each participant receiving one point for each correct placement and one point for each correct verbal label. Points were given in some instances when the verbal answers approximated the area of the body part. There were five areas of verbal approximation that received credit: arm for forearm, head for forehead, finger for index finger, leg for thigh, and throat for neck. Similarly, participants were given points for correct placement of the sticker even if it was not on the exact body part but in the appropriate area (e.g., a participant placed a sticker on his lower calf rather than on his knee). At the conclusion of all 60 trials, participants were thanked for their participation and were reimbursed $2.00 for their time. All participants were then debriefed as to the nature and hypotheses of the study. All information was coded, with identifying information being kept separate from the final coded sheets.
A 2 (level of mental retardation: mild, moderate) × 3 (visual tools type: live models, dolls, drawings) × 2 (response type: verbal, performance) repeated measures multivariate analysis of variance (MANOVA) was used to analyze the results. This three-way interaction approached but did not reach significance, F(2, 31) = 3.02, p = .06. The relative number of correct responses by level of mental retardation is shown in Figure 2. As can be seen, there was little difference between method of responding for individuals with mild mental retardation. Both verbal and performance methods of conveying information appear equally effective, regardless of the visual tools used. The same was not true for those with moderate mental retardation, who were far more able to communicate with behaviors than verbally. This effect was most pronounced for live models when compared to the use of dolls and drawings.
When Visual Tools × Response Type was compared, a robustly significant interaction was obtained, F(2, 31) = 8.47, p = .001. When the live models were used, performance of the behavior was more accurate than were verbal reports for individuals with both mild and moderate mental retardation. When dolls or drawings were used, the difference in accuracy between verbal versus behavioral responses varied less.
The person's level of intellectual ability did significantly interact with the type of visual tools used (Mental Retardation Level × Visual Tools), F(2, 31) = 3.54, p = .035. Clearly, the type of visual tool became more important for individuals with moderate mental retardation than for those with mild mental retardation. Although both groups were more accurate with live models, the difference between live models and the other two methods was greater for individuals with moderate mental retardation. The type of response given also affected accuracy of people with moderate mental retardation more than those with mild mental retardation: Mental Retardation Level × Response Type, F(1, 32) = 5.30, p = .028. Individuals with moderate mental retardation were far more accurate when responding with behaviors than with verbal labeling. Little difference existed for individuals with mild mental retardation. Each individual main effect factor was found to be significantly related to the results. Level of mental retardation was found to be important, F(1, 30) = 9.25, p = .005. The type of visual tool used, F(2, 64) = 10.08, p = .000, and the type of response elicited, F(1, 32) = 6.57, p = .015, were both significantly individually related to the effects found. Two covariants were identified to determine whether these might affect overall accuracy. These covariates were chosen on the a priori assumption that use of psychotropic medications or the existence of a serious mental health diagnosis could significantly impact the person's ability to think clearly. Thus, participants who were receiving psychotropic medications and those who had a secondary mental health diagnosis were individually identified. Neither medication, t = .708, p = .48, nor the existence of another psychiatric condition, t =−1.46, p = .15, was found to affect the results.
Although people with mental retardation have long been the targets of abuse in general and sexual abuse in particular (Furey et al., 1994; Sangrund, Gaines, & Green, 1974; Sobsey, 1994; Westcott, 1993), recent efforts have resulted in more attention being focused on how to correctly identify those who are abused (Valenti-Hein & Schwartz, 1995). In an attempt to find appropriate techniques to use, those in the field of developmental disability have looked to the child abuse literature for inspiration as to methods that would help people with limited cognitive abilities tell their story. Two of these techniques, use of anatomical dolls and anatomical drawings, have been advocated for interviews with adults who have mental retardation and are suspected of being abused. Despite the controversies surrounding the developmental appropriateness of these techniques and the likelihood that they are leading, these methods continued to be used because it is generally assumed that they improved the overall accuracy of reports by those who use them. The cognitive literature on children suggests that the ability to use these tools may be developmentally linked and not just related to developing language regarding identification of body parts. Indeed, it appears as though the generalization skills themselves are what are lacking (DeLoach, 1995; DeLoach, Miller, & Rosengren, 1997; Smith & DeLoach, 1995).
In this study I attempted to replicate this area of children's research by extension to adults with mental retardation. If the cognitive skill of representationalism is an issue for children with more developmental limitations, the same might be true for adults with developmental limitations. Our findings did support this hypothesis in several important ways. First, adults with mild mental retardation were generally superior to those with moderate mental retardation across all tasks. Second, the visual tools used did affect the results with increased accuracy in description occurring when live models versus dolls or drawings were used. This indicates that the generalization skills required for using the visual tools were difficult for all people in this study. Third, all groups showed superiority for showing versus verbal labeling, indicating that demonstrating is important for individuals to fully describe what has occurred.
More important, however, is the consideration of the interactions that occurred. The two-way interaction between level of mental retardation and type of response indicated that demonstration was particularly superior for people with moderate mental retardation compared to verbal methods of conveying the body part being identified. This suggests that people with moderate mental retardation should be given a means during interview settings by which they can demonstrate what happened to them. An interaction also occurred between the visual tools used and the level of mental retardation of the participant. Thus, choice of visual tools becomes more critical for adults with moderate mental retardation who showed significant drops in ability to convey information when dolls and drawings were used. Some case studies with young children have shown that when given anatomical dolls to interact with in an interview situation, many will use the doll against their own body to show what happened (M.D. Everson, personal communications, March 1996). I have seen similar results in clinical settings when adults with mental retardation are given anatomical dolls to convey their stories. Such demonstrations using themselves as part of the story might be considered to have added credibility.
An interaction also occurred between the type of visual tools used and the response type given. When live models were used, demonstrations were far more accurate than were the verbal reports for all participants. This occurred even though the live model was female and might have increased the challenge of generalization for male participants. Because the overall level of accuracy declined with the use of dolls and drawings, the difference between verbal versus demonstration in accuracy appeared less critical. Thus, although dolls are given to improve descriptive accuracy by providing the individual a way of showing what happened, it appears as though their utility is more in the realm of what Everson and Boat (1994) referred to as a memory aide: They jog the individual's memory such that they are able to convey more information when this information is conveyed in a verbal means.
Although results of this study suggest that the use of visual tools might not be as beneficial as originally thought, this research should not be construed as reason to disregard their use altogether. Indeed, this study shows the unique role of these methods as a memory jogger for adults with mental retardation. If the participant has mild mental retardation, the tools hold particular value, and verbal or behavioral use of them may add to the reporting ability. If the participant has moderate mental retardation, the tools may hold less value, but add a communication tool. Even more validity might be granted if the person could demonstrate by use of his or her own body as a live model (e.g., “Point to the part on your body that was touched”), either with or without the use of the dolls.
In this study I only examined the ability of individuals with mental retardation to use visual tools to represent themselves. An equally important issue is whether use of visual tools helps to elicit a more complete relating of events without losing accuracy. If the visual tools can be used by an adult with mental retardation, the question remains as to whether the use of such visual tools adds to the completeness of the story. A very important future direction for research is the use of visual tools to relate elements of a contrived event with known elements so that the accuracy and additive elements of a multimodal approach to interviewing might be assessed.
These findings have implications not only for forensic practices, but for education. Program instruction on sexuality and self-protection skills come to mind. Often in these programs, dolls and drawings are used to teach concepts. Indeed, for some people, training might be better understood without the use of representations if the learners could directly apply the information to their own bodies. So much of self-protective programming is focused on assertive refusal, boundaries, labeling of feelings, and avoiding risky situations. The results of the present study suggest that we may need to take a step back and focus more on identification of body parts and the ability to relate personally experienced events as basic skills for self-protection.
This study presents an initial evaluation of the ability of adults with mental retardation to use visual tools for telling their story. Clearly, some concerns might be leveled, including the sample size. Further, levels of functioning were based on recorded IQs for determining experimental groupings. One may argue that IQ is not the most important variable in determining the ability to use visual tools and that other measures of individual differences (such as adaptive behavior levels) might better predict outcome on this task.
Sexual abuse has many psychological consequences (e.g., withdrawal, noncompliance, and aggression) that could further affect the ability to use these visual tools (Sobsey, 1994). Although results of this study indicate that adults are able to use these methods in a laboratory situation, they may not reflect the actual typography of responding when the individual is under the stress of conveying what happened to them after a traumatic situation in the unfamiliar setting of a police station or clinic therapy room. Future research attempts to replicate conditions closer to those that exist in actual abuse situations are needed.
Note: The author thanks the National ARC for funding this research and The Milwaukee Center for Independence for providing space and soliciting participants.
Protocol for Cognitive Assessment Project—Body Identification
Solicit participation from agency consumers via staff.
Read consent form to consumer and have them sign. Get guardian signatures when appropriate.
Read directions to the study: “We are trying to find out what types of body parts people can find on themselves and on other people. I will put this dot on my body. I want you to put this dot on the same part of your body. Please name that body part while you are doing this.”
Have participants identify the following body parts by placing the dot on your body out of view of the individual: (1) cheek, (2) index finger-right, (3) elbow-left, (4) stomach, (5) back, (6) right eye, (7) right foot, (8) left knee, (9) lips, (10) buttocks, (11) left ear, (12) right forearm, (13) neck, (14) lower back, (15) left thigh, (16) forehead, (17) hair, (18) bottom of one foot, (19) right shoulder, (20) left breast/chest area,
Read these directions: “Now I want to do this again. This time, I am going to use this doll. I will put the dot on the doll's body part and I want you to put it on the same part of your body. Please name the body part when you do this.
Place dot on the doll's body out of view of the individual. The order used is backwards. (Present doll with Body Part (20)—left breast/chest marked first.)
Read the following instructions: “You're doing great! We are going to do this one more time. This time, I am going to use this drawing to place the dot on the body. The drawing shows a person without clothes, but you can place this dot on top of your clothes. Please name the body part as you are placing the dot.”
Place the dot on the following body parts out of sight of the individual.
Thank them for participation, debrief, and pay them $2.00.
Use this order for the first 10 clients. For the second 10 clients, begin with the dolls. For the third 10 clients, begin using the drawings. Repeat with clients who are lower functioning.
Author:Denise Valenti-Hein, PhD, Adjunct Professor, Department of Psychology, 800 Algoma Blvd., University of Wisconsin-Oshkosh, Oshkosh, WI 54901–8670