Assessment

Do the research trials investigating the potential effectiveness of the rehabilitation method meet at least minimal professional standards? For example, is the research conducted independently? (We are rarely disinterested judges of the profundity, effectiveness, and near-perfection of our own work.)

A more complex requirement concerns whether the base rate of discovery of abuse is adequately taken into account in conducting and reporting the results of experimental trials of rehabilitation efforts. Perpetrators may continue to engage in sexual intimacies with clients during (or after) rehabilitation efforts, even when they are supervised (see, for example, Bates & Brodsky, 1989). The abuse may come to the light only if the client reports it. Yet the base rate of such reports by clients is quite low. Surveys of victims suggest that only about 5 per cent report the behavior to a licensing board (see Pope & Vetter, 1991). The percentage appears to be significantly lower when the number of instances of abuse estimated from anonymous surveys of clinicians (who report instances in which they have engaged in abuse) is compared with complaints filed with licensing boards, ethics committees, and the civil and criminal courts. Using the higher 5 percent reporting estimate, assume that you conduct research in which a licensing board refers ten offenders to you for rehabilitation. You work with the offenders for several years and are convinced that you have completely rehabilitated all ten. You assure the licensing board of your complete confidence that none of the ten will pose any risk to future clients. But also assume that your rehabilitation effort fails miserably: all ten offenders will engage in sex with a future client. What are the probabilities that any of the ten future abuse victims will file a complaint? If each client has only a 5 percent probability of reporting the abuse, there is a 59.9 percent probability that none of the ten will file a complaint. Thus, there is close to a 60 percent chance that these research trials, even if independently evaluated, will appear to validate your approach as 100 percent effective when in fact it was 100 percent ineffective. If ignored in conducting and reporting research, the low base rate can make a worthless intervention appear completely reliable.

0 0

Post a comment