154206 Measuring impacts without footprints: An evaluation of Ghana's AYA Project

Monday, November 5, 2007: 3:00 PM

Michael McQuestion, PhD , Consultant, CHIME Project, John Snow, Incorporated, Baltimore, MD
Ali Mehryar Karim, PhD , Last 10 Kilometers Project, JSI Research & Training Institute, Inc., Addis Ababa, Ethiopia
Jessica Posner, MPH , John Snow, Inc., Arlington, VA
Timothy Williams, MA, MEM , John Snow, Inc., Arlington, VA
Clement Ahiadeke, PhD , Institute for Statitical, Social and Economic Research, University of Ghana, Accra, Ghana
Public health programs often overlap or do not “brand” themselves, making it hard to distinguish exposures. Interventions may even be associated by design such that one is meant to enhance another, leaving no footprint. This was the case with the African Youth Alliance, a five-year adolescent sexual and reproductive health (ASRH) program carried out in four African countries during 2002-2005. In this paper we present our evaluation findings from one AYA country, Ghana. We used administrative and census data to draw a purposive sample of 105 treatment area clusters and 75 comparison group clusters. We surveyed 3,362 youths ages 17-22. Our instrument contained eight exposure items, both general and specific to the AYA/Ghana Project. Another battery of items probed respondents' knowledge across seven key ASRH domains. We cross-tabulated the exposure and ASRH content items to create two indexes: one for AYA/Ghana exposure and another for background ASRH exposure. We dichotomized the indexes and modeled them as potentially endogenous predictors of four behaviors: abstinence, modern contraceptive use at first sex, condom use with last partner and two or more sexual partners in the preceding 12 months. We modeled males and females separately using recursive trivariate probits. Estimation was by simulated maximum likelihood. All models included exclusion restrictions and controls for socioeconomic status, household structure, region and other factors. Post estimation probabilities from these eight models show that salutary outcomes were most likely among youths exposed both to AYA and to other ASRH interventions. AYA's treatment effects were consistent only among females.

Learning Objectives:
Participants will see how one can differentiate treatment effects from related programs using a multivariate probit model.

Keywords: Sexual Risk Behavior, Evaluation

Presenting author's disclosure statement:

Any relevant financial relationships? No
Any institutionally-contracted trials related to this submission?

I agree to comply with the American Public Health Association Conflict of Interest and Commercial Support Guidelines, and to disclose to the participants any off-label or experimental uses of a commercial product or service discussed in my presentation.