Sexual Abuse: A Journal of Research and Treatment [saj] PL101-170 March 14, 2000 9:39 Style ﬁle version Nov. 19th, 1999
Sexual Abuse: A Journal of Research and Treatment, Vol. 12, No. 2, 2000
Problems with the DSM-IV Diagnosis of Pedophilia
Lisa G. Regev,
and Anne Hagstrom
This paper examines the taxonomic adequacy of the Diagnostic and Statistical
Manual, 4th ed., DSM-IV (American Psychiatric Association, 1994) diagnostic
category of pedophilia. This diagnosis, as well as the other sexual disorders,
have been ignored in DSM ﬁeld trials. There is no empirical information about
the reliability or validity of this diagnosis. Moreover, because the vagueness of the
diagnostic criteria, clinicians would need to make inferences that would likely lead
to reliability problems in diagnosis. Further, the DSM diagnostic criteria include
constructs that are not intersubjectively veriﬁable and for which there are no valid
measures. This can also lead to lack of diagnostic reliability and accuracy. Most
problematical however, there are aspects of the diagnostic criteria, most notably
the presence of an “ego dystonic sexual attraction to children,” that are incorrect
exclusion criteria. Suggestions for improvement are provided.
KEY WORDS: pedophilia; sexual interest; DSM-IV; behavioral disorder.
Any lexical or any more formalized deﬁnition of pedophilia (from the Greek,
ioral indices or manifestations of this sexual interest can include sexual arousal to
children, sexual fantasies involving children, and acts such as buying child pornog-
raphy or touching a child in a sexual way. Sexual interest is notoriously difﬁcult
to measure accurately because some of its key manifestations, such as fantasies,
sensations of attraction, and urges, are covert. Moreover, given the tendencies of
child molesters to deny or minimize their deviant sexual interests (Abel, Becker,
Mittelman, Cunningham-Rathner, & Rouleau, 1988; O’Donohue & Letourneau,
1993) self-report or self-report-based measuring instruments that attempt to mea-
sure overt behavior also have, at best, questionable accuracy.
University of Nevada, Reno.
Northern Illinois University.
2000 Plenum Publishing Corporation