ICAR Research Data Repository for Knowledge Management

The current research study was designed to provide quantitative data regarding the sexual networks of married and single RDF personnel, to measure levels of condom use of RDF members by partner type and to explore associated determinants of condom use, including, but not limited to, exposure to HIV prevention programmatic interventions and alcohol-related behavior. The research study also explores attitudes toward œgender-equitable norms in the military. Results will be used to inform
program targeting and planning.This study is a nationally representative study of military personnel in the RDF. A total of 2000 military members were interviewed. The final usable sample size was 1956; data from 44 interviews were unusable due to incomplete interviews or corrupted data. The study was initiated after receiving all appropriate written authorizations. Approval was received from the National Ethics Committee in April of 2009 and from DMS in June of 2009.

A two stage sampling approach was used for sample selection. At the first stage, the sampling frame consisted of the eleven RDF brigages and three small military schools in the country. Because of their relatively small size, the military schools were considered as a single sampling unit; if selected, all were sampled. A random sample of eight of the twelve RDF brigades and/or military school was selected at the first stage. The following brigades were selected: Nyagatare, Nyamagabe,
Muhanga, Huye, Karongi, Gicumbi, Musanze, as well as the military schools (Gabiro, Nasho and Gako). Since the size of the brigades was not known in advance, a constant number of 250 military personnel was selected per brigade at the second stage and a total of 250 RDF were selected from the three schools. Due to the sensitive nature of the information involved in conducting sampling, PSI/Rwanda trained DMS staff and members from each selected brigade to carry out this task. A PSI/Rwanda staff member was present to respond to questions but did not have access to any identifying information pertaining to RDF personnel.

The proportion of officers, junior officers and soldiers were estimated in each brigade prior to conducting the sampling. These proportions were then applied to the 200 RDF sampled in each brigade to ensure that the sample was representative by rank. More detailed methodological information is available in the study design document upon request to PSI/Rwanda.