Questionnaire and interview inconsistencies exaggerated differences between adopted and non-adopted adolescents in a national sample

In a series of adoption-related studies using the Add Health data sets, discrepancies were observed between the Add Health In-school Questionnaire and the In-home Interview data sets in the com-Address correspondence to: Xitao Fan, Curry School of Education, University of Virginia, 405 Emmet Street South, Charlottesville, VA 22903 (E-mail: [email protected]). parisons between adoptees and non-adoptees on a group of behavioral and psychosocial outcome variables. Detailed analyses led to our identification of a group of adolescents who were likely to have mischievously or dishonestly identified themselves as “adoptees” when in fact they were not. This group also appeared to have provided extreme responses on questions about delinquent behaviors. The evidence suggests that, as a result of the extreme responses of this group, the previously reported differences between the adopted and non-adopted groups in our article in Child Development(Miller et al., 2000) may have been substantially overstated. Substantive implications of these findings are discussed.