What Were They Thinking? An Experimental Investigation of Child Sexual Offenders' Beliefs.

Abstract

In the field of forensic psychology, child sexual offenders (CSOs) are often hypothesised to hold abnormal beliefs that facilitate the onset and maintenance of their offending. This idea has had considerable impact upon current CSO assessment and treatment practices. However, despite its intuitive appeal, empirical evidence supporting the hypothesis is unfortunately lacking. Information regarding the role that cognition plays in child sexual offending has been gathered almost exclusively using self report (i.e. interview and questionnaire) methods. In interview studies, CSOs talk at length about their offending and their statements are analysed for the presence of so-called cognitive distortions: utterances deemed to represent abnormal, offence-facilitating beliefs. In questionnaire studies CSOs and controls rate the veracity of listed cognitive distortion items and their answers are compared. In general, interview and questionnaire studies have tended to find that CSOs endorse cognitive distortions, which seemingly supports the notion that they hold offence-supportive beliefs. However, serious issues plague the use of these self-report methods because endorsement of cognitive distortions might reflect phenomena other than beliefs. The primary aim of this thesis was to examine the idea that CSOs hold offence-supportive beliefs using methods designed to side-step issues associated with self-report methods. Across three studies, three cognitive experimental techniques were for the first time applied to the study of CSO cognition. In Study One, CSOs and offender and community controls completed an experimental procedure called the lexical decision task. Against hypotheses, when compared to controls CSOs did not interpret offence-related sentences in line with distorted beliefs. A possible explanation for this finding was that CSOs' offence-supportive beliefs were insufficiently activated during testing. To investigate, in Study Two half the CSO and half the offender control participants were primed with images of scantily-clad children before commencing experimental testing. During testing, CSOs and offender controls read sentences describing children behaving in potentially sexualised ways. Participants were then given a surprise recognition test in which half the sentences were re-presented in an unambiguously sexual form, and half in an unambiguously nonsexual form. Contrary to hypotheses, neither primed nor control child sexual offenders showed memory biases for sexualised sentences, suggesting they did not interpret the original sentences in line with offence-supportive beliefs. Finally, in Study Three, CSOs' beliefs were examined using interview methods, and CSOs' and offender controls' beliefs were measured using a questionnaire as well as an experimental technique that used sentence reading times to implicitly measure beliefs. As hypothesised, CSOs showed evidence of holding offence-supportive beliefs according to the interview and questionnaire measures, but against predictions they demonstrated no experimental evidence of such beliefs. In fact, the three methods showed virtually no agreement regarding the belief-types each CSO was deemed to hold, raising important questions about the phenomena measured by each method. Overall, the results of this thesis did not support the stance that CSOs generally hold offence-supportive beliefs that set them apart from others. The implications of these findings for theory and treatment are discussed and directions for future research are suggested.