An Exploratory Study about Inaccuracy and Invalidity in Adolescent Self-Report Surveys

Abstract

Using Add Health data, the authors provide evidence that some adolescents gave inaccurate and/or invalid responses on a self-administered questionnaire. Further analyses show that these adolescents were much more likely to report extreme levels on psychosocial and behavioral outcome variables. A distinction was made between inaccurate responders (e.g., inaccurate/false responses due to carelessness or confusion) and jokesters (e.g., intentional false responses). The findings show that the jokesters showed considerably more pronounced distorting effects on some psychosocial and behavioral outcome variables than the inaccurate responders did. The authors suggest that although this jokester effect may not seriously bias the results in studies that focus on large groups, for research focusing on some special subgroups (e.g., adoption groups, immigrant groups, disability groups), this effect could pose a serious challenge for the validity of research findings.