|Learning to Live - Monitoring and Evaluating HIV/AIDS Programmes for Young People - Practice Handbook (DFID - UNAIDS - Save the Children, 2000, 220 p.)|
|Chapter 9: Data Collection: Where do we get our information and how?|
Combining methods and indicators in a single M&E framework can show: changes within the target group; the relationship between changes and interventions. Asking respondents where they receive their information, and then looking at the accuracy of that information (or associated behavioural indicators), can help assess the impact of the intervention, An example from Zimbabwe shows how (Box 9.4).
Box 9.4 Determining attribution in community-based interventions: the case of an outreach project in Zimbabwe
The HIV/AIDS Prevention in Africa (HAPA) Support Program (HSP) at the John Hopkins University (JHU) provides technical support to NGOs in the field. Save the Children US (SC US) requested assistance to evaluate projects in Zimbabwe, Baseline data had been collected at the start of the project, which was in 3 sites - each with approximately 12,000 population. The aim of the project was to train Ministry of Health and SC UK staff to provide outreach community education. For the evaluation, a 32-item questionnaire was prepared, with a team of 10 supervisors and 42 interviewers. The total sample size in 2 sites was 660 persons, aged 18-45. A 30-cluster sample method was used, modified from the technique developed by the WHO for immunisation coverage surveys. There were 3-4 days for data collection.
85 per cent of respondents were able to recognise the 3 main modes of HIV transmission (project objective had been 80 per cent), but belief in mosquito transmission was high. Only 30 per cent of respondents thought children under the age of II should be educated about AIDS but 80 per cent thought that 11-15 year olds should be. Interestingly in baseline, more people thought there was no cure for AIDS than in the follow-up survey Follow-up data showed higher awareness of the value of condoms, and faithfulness.
HIV/AIDS-prevention efforts by other groups were not taking place in the SC UK impact areas during the period of work, so no project evaluation was possible using a comparison group. But the Minister of Health did raise awareness in a national campaign.
For attribution testing, respondents were asked about their sources of information. Respondents could be categorised into 2 groups:
(i) Those mentioning a group targeted by SC UK for training (e.g., community health workers or community leaders) as at least one information source.
This gave information on gender coverage (more women than men mentioned an SC-trained source), and also the level of coverage in the various sites.
1. Significantly more respondents naming SC UK sources of information (compared to those listing other sources) could spontaneously list the 3 main modes of HIV transmission.
Overall, the survey findings reinforced the need for HIV/AIDS-education efforts to target changes in attitudes and behaviour rather than focusing primarily on increases in knowledge, and to use multiple strategies to accomplish that end.
Source: Mercer and others, 1996.