A treatment group, in an experiment or impact evaluation, is the group of participants who receive the intervention being studied, such as a loan offer, a savings account, a training course or a cash transfer. Its outcomes are compared with those of a control group that does not receive the intervention, and the difference is taken as an estimate of the programme's effect, a step known as attribution.
The comparison is most reliable when people are assigned to the two groups at random, as in a randomized controlled trial, because randomisation makes the groups alike on average in everything except the treatment. Where random assignment is not possible, evaluators use statistical methods to build a comparable group. In microcredit research, several randomised studies in different countries, six published together in 2015, compared areas where credit was introduced with similar areas where it was not, and found modest average effects on household income. Such studies are central to impact assessment.
Ref 1270000024