Anatomically detailed dolls are sold by different manufacturers; and as a result, the various dolls are not always comparable to each other. Some of the dolls have oral, anal and vaginal openings, but others do not. Moreover, there is no generally recognized and accepted method for using anatomically detailed dolls when interviewing children. For example, some professionals video-tape children’s interactions with the dolls, but others do not. How children respond to the dolls can be as influenced by the particular doll used, or how the child is interviewed, as by whether or not the child has been abused. Not surprisingly, then, different professionals can reach different conclusions relying on anatomically detailed dolls when interviewing the same child (Boat & Everson, 1988).
A 1989 study systematically recorded how 91 non-abused children, ages 2-6, responded to anatomically detailed dolls (Glaser & Collins, 1989). With little to no encouragement, 74% of these children spontaneously undressed the dolls. Trained observers concluded that 64% of the children exhibited various kinds of emotional reactions to the dolls while undressing them. 71% of the children touched the anatomically detailed doll’s penis, 13% touched the anus, and 4% touched the vaginal opening (Glaser & Collins, 1989).
Interestingly enough, many professionals contend that spontaneously undressing an anatomically detailed doll is consistent with children who have been sexually abused (Myers, Saywitz, & Goodman, 1996). Other professionals frequently conclude that children have been sexually abused as a result of their exhibiting an emotional reaction to the dolls. Furthermore, these professionals would consider it quite likely that children have been abused if they touch the dolls’ genitalia. Given these kinds of decision making criteria related to anatomically detailed dolls, it is possible to mistakenly conclude that as many as 74% of non-abused children have been sexually abused. To say the least, this is a rather shocking rate of error.
Another study examined how 35 children between the ages of 2 and 6, referred for evaluation of possible sexual abuse, interacted with anatomically detailed dolls. The responses of these children were compared with another 35 children of the same age who had not been sexually abused. Of the children referred for possible sexual abuse, 30 of them touched the anatomically detailed dolls’ genitalia at least once during the interview, but 25 of the non-referred children also did the same. Nine of the referred children engaged in sexually explicit play with the dolls; however, five of the non-referred children also responded in this manner. In other words, this study found no evidence indicating that anatomically detailed dolls can reliably differentiate between children referred for sexual abuse evaluations, and children who have not been abused.
In view of the evidence cited above, it is not surprising that anatomically detailed dolls have fallen into disrepute (Bruck, Ceci, & Francoeur, 2000; Ceci & Bruck, 1995; Pipe & Salmon, 2009; Poole, Bruck, & Pipe, 2011). Relying on anatomically detailed dolls can lead professionals into two kinds of errors: (1) Mistakenly classifying children who have been sexually abused as not abused, and (2) Mistakenly classifying children who have not been sexually abused as abused (Bruck, Ceci, Francouer, & Renick, 1995). The frequency with which anatomically detailed dolls has led to these kinds of errors has resulted in one reviewer emphasizing: “Anatomically detailed dolls are not useful, and should not be used, for determining whether or not sexual abuse has occurred” (Maan, 1991, p. 49).
Various professionals continue to defend the use of anatomically detailed dolls despite the issues and evidence cited above (Faller, 2007; Hlavka, Olinger, & Lashley, 2010). These professionals insist that anatomically detailed doll interviews are only one component of a more comprehensive evaluation. Their reasoning suggests that even though anatomically detailed dolls result in an unacceptable number of mistaken classifications, using the dolls along with other evaluation procedures is somehow justifiable. Such thinking is tantamount to claiming that when combined with sophisticated procedures such as CAT scans, phrenology – measuring the various bumps and curvatures of the head – can effectively assist neurologists in their diagnostic endeavors. While emphatically rejecting this kind of nonsensical argument, other reviewers have also deplored the use of anatomically detailed dolls: “We are left with the conclusion that there is simply no scientific evidence available that would justify clinical or forensic diagnosis of abuse on the basis of doll play” (Wolfner, Faust, & Dawes, 1993, p. 9).