Methodologic Note for Child Epidemiological Surveys: The Effects of Instructions on Estimates of Behavior Prevalence

This paper examines the effects of a change in the wording of the Child Behavior Checklist (CBCL) instructions for the 0 category on the reported prevalence of behaviors among normal children by their parents. Convenience samples of parents were randomly allocated to complete CBCL items using either the standard instructions "not true" or "never or not true" to describe the 0 category. Adding the word "never" to describe the 0 category decreased parents' use of that category and increased their use of the adjacent, 1, category. Implications of these findings are discussed.