To establish cause-effect relations, a logic model based on change theory must be developed, which explains the action of the programme and the expected effect on the population. It forms the basis for the causal chain which we assume occurs once the project has started; this relation between causes and expected effects is what we will call the hypothesis. Very occasionally, however, an effect is caused by just one cause, while there are other factors that influence the relation between the actions and effects of the project; contextual variables which limit or accelerate this phenomenon are called intervention variables.

To construct hypotheses, the causal relation between the action and expected effects of the programme should be considered, in addition to specific aspects which may affect the expected outcome. For example, as regards access to credit, the numbers should be determined not only of those who received a new title who were given credit, but also of people applying for credit, how many met the bank’s requirements and what proportion of those obtained it. It is consequently important to establish appropriate hypotheses for each of these aspects, defining the direct variables of the intervention and those affecting how the outcomes are obtained.

See Fact sheet for designing and processing household surveys for further details on the hypotheses.