Doctor of Philosophy (PhD)
Child assessments have consistently yielded poor agreement between parent-child self-report measures and considerable research has failed to isolate the underlying source of the discrepancies. Validity testing has successfully been used to differentiate credible from non-credible performance, recognize feigned symptoms, and detect underestimates of ability in adults for decades (Sherman, 2015). Despite these potential benefits, attention to the importance of validity testing in the child assessment literature has been largely neglected. Many psychologists believe children are not capable of feigning or exaggerating their symptoms and lack the underlying motivations and incentives for deception found in adult assessments. Whereas, more research is needed to determine the legitimacy of these beliefs, the reliance on parental reports when assessing child symptoms provides an opportunity to integrate symptom validity testing into child assessment practices by using a population where it has already been proven effective. The current study utilized polynomial regression analyses in order to determine whether an embedded symptom validity measure could predict symptom agreement across informants. Participants included 220 children (42.7% female) ages 11-17 (M = 12.52, SD = 1.43) and their parents. Four separate polynomial regression analyses were conducted across symptom constructs for which children are commonly referred for outpatient treatment services (depression, anxiety, inattention, and aggression). Results showed that parental desirable responding was detected across several constructs and predicted informant discrepancies in anxiety and aggression symptoms.
Lilly, Megan Elizabeth, "Polynomial Regression Analyses of Parental Desirable Responding" (2017). LSU Doctoral Dissertations. 4176.
Davis III, Thompson