Guidance for the Design of Quantitative Survey-Based Evaluation

Survey-based evaluation aims at surveying respondents from a statistically valid sample of the population of interest in order to infer from the sample statistics the range of likely values in the broader population. The population of reference can be made up of very different elements. It can be credit institutions and borrowers in policies related to credit, it can be farmers in farming policies or the entire group of SMEs in a region for policies target to this economic context. The relevant population from which to sample is defined by the nature of the policy intervention one wants to examine. In what follows I will use the term ―agents‖ to define the elements of the target population. Survey-based quantitative evaluations are frequently used to assess how people/agents feel about the realization of policy outcomes before, during and after a given policy intervention. The term ―quantitative‖ refers to the fact that there is normally a numerical scale behind the quantities that respondents are asked to reveal in the survey. This numerical scale can be as simple as expressing a degree of agreement or disagreement over a scale of ordinal intensities.