Major points to remember regarding data collection

· Proof of impact of an intervention will not be demonstrated, and nor should a project attempt to demonstrate it.

· Mix your methods - an ideal evaluation will have qualitative and quantitative data.

· No one method is perfect: remember this when analysing data. Analysis should be thematic rather than by method, i.e., when looking at condom use, evidence from different methods used (focus groups, informant interview, PLA techniques, questionnaire, community-based distributor data) should be considered together.

· The general rule regarding questionnaires is that the longer the interaction time between interviewer and respondent, the more honest and reliable the answers.

· Focus group discussions can provide both quantitative and qualitative data.

· Anonymity and confidentiality are essential when discussing sensitive issues.

· Clinical records are a vital source of information.

· Young people assisting in the design of research tools, and collection of information, is empowering to the project, and helps to ensure relevance of the information provided.