Pitfalls in adapting general population measures for use in research with people with intellectual disabilities

Abstract

Objectives:In a study concerning empathy in offenders who have intellectual disabilities (ID) we adapted a measure designed for use with the general population. In this paper we describe how we did this and the pitfalls we encountered. We were looking at empathy because it is seen as a significant factor in treatment and treatment outcomes with sex offenders. This has been inducted into treatment programmes and evaluations with sex offenders who have ID. The aim of our study was to explore general and victim empathy in sex-offenders and non-offenders with ID. The original aim was to explore differences between these groups on adapted measures of general and victim empathy, and if there was any relationship between the sex-offenders’ scores on empathy towards non-specific victims of sexual assault and their empathy towards their own victim.Methods:A measure of general and victim specific empathy developed for use with sex offenders in the general population (the Empat) was adapted for use with adults who have ID. Twenty-one sex-offenders and 21 non-offenders with ID, matched by age, gender and IQ, completed the adapted version of the Empat and the offender group also completed a previously adapted version of the Victim Empathy Scale.Results:Psychometric evaluation of the adapted scale produced marginal and poor reliabilities. In view of this further analysis to examine comparisons between the two groups would most likely produce spurious results. Thus the introduction of measures from the general population to populations of people who have ID are not always successful. However, such outcomes are not always reported. Further research is needed to develop better approaches to the assessment of victim specific and general empathy in offenders who have ID.Conclusion:Despite following guidelines on measure adaptation for people who have ID, the measures did not achieve satisfactory psychometric properties and, therefore, any findings could be spurious. This study clearly shows that researchers need to establish that any general population measures they employ or adapt for use with people who have ID need to be evaluated psychometrically first.