Measuring change in behaviour : an analysis of a participatory evaluation method in hygiene education for water and sanitation

This thesis evaluates the pilot phase of a hygiene education programme in Boucle du Mouhoun, a region in north west Burkina Faso, during which intervention took place in 59 villages. It analyzes the programme 's effect evaluation with respect to: the evaluation design, the relevancy of the chosen messages, the validity of the indicators and their selection procedures, the data collection methods (including reliability of the instrument, bias, selection and training of the observers and the data collection procedures), data analysis and participation. Certain aspects of the process evaluation are briefly reviewed as well. The main findings and recommendations of this analysis are:
- The evaluation design was relatively effective, although no control group was used. It took on average 10% of the time spent by the actors on hygiene education, which is acceptable.
- The messages addressed the public health problems in the area, but the specific situation in the villages should be better taken into account, for instance with respect to issues such as the use of traditional wells.
- Indicators were chosen on the possibility to observe them and some behaviours were therefore not included in the evaluation. For this reason results could not be generalized.
- Structured observations were used for data collection. However, it appeared to be difficult to develop simple data collection procedures.
- Data analysis was carried out at regional level, although discussions of the data took place with all the actors. Results were presented per province and per village.
- Although participation of the health staff and the community in the effect evaluation is required at all stages, it needs to be more operationalized in the evaluation procedures and in data analysis.
- the aspects of the process evaluation that were reviewed in this thesis were participatory and included all the actors of the programme throughout. However, the underlying theoretical concept of peers influencing behavioural change needs to be evaluated.
In general, the evaluation system was carried out on a continuous basis using a clearly defined system. There are sufficient opportunities to address the weaknesses mentioned above, in order to improve future evaluations. (author's abstract)