Publication

Abstract : Amrita Awareness Ambassador Program is a substance-abuse prevention program that aims to empower local youth in rural India. The program empowers and promotes peer mentors from local schools to act as awareness ambassadors. The program is already active in 13 states. To adapt the program to new villages, substance abuse prevalence in the considered region has to be assessed first. One difficulty of such an initial assessment, especially in tribal areas, is that villagers often answer questions about substance abuse in an inconsistent and unreliable way when talking to outsiders. This paper reports on findings from a remote, tribal area in Chhattisgarh, where 132 adolescents and 11 awareness ambassadors were asked about the prevalence of abuse of alcohol and other substances. Four types of statistical consistency are considered in the analysis: internal, external, individual, and dimensionality consistency. Statements from the awareness ambassadors are found to be significantly more internally and externally consistent and reliable than statements from the village adolescents. In addition, statements from the awareness ambassadors also show high inter-rater and dimensionality consistency. As an implication of these findings, we conclude that while planning a substance abuse program, reliability of the information sources needs to be taken into explicit consideration.