Statistical non-significance vs. practical relevance in intervention evaluation of sexuality education programs

Abstract

Measuring effectiveness of sexuality education for high-risk target populations using quantitative measures is difficult. The lack of significant results erroneously indicate a lack of effectiveness. Reasons for lack of significant results are paramount, and include participants' trouble with self-report questionnaires, scarcity of risk-behavior, and a difference between intervention-as-intended and intervention-as-realized. In an attempt to supplement promising (yet non-significant) results of a quantitative evaluation, in-depth interviews were held with 13 participants of an group counseling intervention for girls with high-risk for sexual abuse, disease, and forced prostitution ('loverboys'). Participants who joined the group counseling voluntarily differed from those who participated involuntarily in reported effects on knowledge, intention and behavior, in that voluntary participants attributed changes in determinants and behavior to the group sessions, whereas involuntary participants did not. Involuntary participants did report changes, however. With the results from the interviews, the effectiveness of the intervention can be better valued than with the quantitative results alone. This adds to the discussion of the use of qualitative research in effectiveness evaluation studies.