Poster Session #1: UC Ballroom
Catching Abusers When They Lie
Presentation Type
Poster
Faculty Mentor’s Full Name
Paul Silverman
Faculty Mentor’s Department
Psychology
Abstract / Artist's Statement
In trying to detect adult child abusers, psychologists use various measures, whose validities are unfortunately limited by the social desirability bias. Respondents are motivated to appear better parents than they really are, especially in child abuse cases. This bias threatens the validity of a widely-used measure, the Child Abuse Potential Inventory (CAP), and another measure, the Parenting as Social Context Questionnaire (PASCQ). Faking good on both the CAP and the PASCQ is quite feasible because of high face validity (it is obvious what each question is measuring). To address this weakness, the Child Guidance Inventory (CGInv) presents specific child guidance scenarios whose most desirable response is ambiguous. To compare the strength of the CGInv against both existing measures, 160 psychology students were administered the measures and given instructions on how to respond, either to answer honestly or to fake good. Half of the sample completed existing measures and the other half completed the CGInv. The analysis compared differences in scores between instruction conditions. It was hypothesized that, despite the instruction condition, participants administered the CGInv would generate more similar scores than would participants administered the existing measures. Groups instructed to fake good on the CAP and PASCQ produced lower maladaptive scores than groups who answer honestly. In contrast, both groups generated similar scores on the CGInv. Given the success of the CGInv to prevent successfully faking good, research should be conducted to further develop this instrument. It may be useful in clinical settings as a convenient, yet valid, self-report measure.
Catching Abusers When They Lie
UC Ballroom
In trying to detect adult child abusers, psychologists use various measures, whose validities are unfortunately limited by the social desirability bias. Respondents are motivated to appear better parents than they really are, especially in child abuse cases. This bias threatens the validity of a widely-used measure, the Child Abuse Potential Inventory (CAP), and another measure, the Parenting as Social Context Questionnaire (PASCQ). Faking good on both the CAP and the PASCQ is quite feasible because of high face validity (it is obvious what each question is measuring). To address this weakness, the Child Guidance Inventory (CGInv) presents specific child guidance scenarios whose most desirable response is ambiguous. To compare the strength of the CGInv against both existing measures, 160 psychology students were administered the measures and given instructions on how to respond, either to answer honestly or to fake good. Half of the sample completed existing measures and the other half completed the CGInv. The analysis compared differences in scores between instruction conditions. It was hypothesized that, despite the instruction condition, participants administered the CGInv would generate more similar scores than would participants administered the existing measures. Groups instructed to fake good on the CAP and PASCQ produced lower maladaptive scores than groups who answer honestly. In contrast, both groups generated similar scores on the CGInv. Given the success of the CGInv to prevent successfully faking good, research should be conducted to further develop this instrument. It may be useful in clinical settings as a convenient, yet valid, self-report measure.