Findings, implications and lessons learned

This case study has demonstrated various difficulties that can arise when attempting an impact assessment study.

• The length that impact assessment studies can take should not be underestimated, especially when the study is in the form of a 'before/after' study
• Before embarking on such a study it is important to make sure that there are sufficient resources to complete the study in the time planned.
• Sometimes, as here, it may be impossible to identify and utilise a set of parallel control subjects (e.g. herds or farmers) to which the intervention is not applied.
• There are potential problems, especially in 'before/after' studies, in distinguishing the impact of an intervention from other confounding factors
• Statistical analysis can be complicated unless data are summarised over longer observational periods (in this case six months) than the frequency at which data are collected (in this case one month).
• Data collection can get out of hand unless the project has strong data management support and there are well-designed data collection systems with appropriate data checking routines in place.