RANDOMIZED CONTROLLED TRIALS
Evaluations based on experimental designs employ randomized controlled trials and provide highly reliable evidence of effectiveness. These counterfactual evaluations represent the gold standard in evaluation. For this reason they are frequently used to assess the effectiveness of medical interventions and have also been used in education, labour training, and international development. See TEN's Margaret Dalziel discuss the findings from a recently conducted randomized controlled trial here.
Governments that want irrefutable evidence of the impact of their investments in support of innovation, entrepreneurship, and research may wish to design their interventions in such a way as to allow for experiment-based evaluation.
This may be possible where programs are large and many firms are eligible for support. Randomized support may be offered to a subset of eligible firms, with the balance of eligible firms receiving non-randomized support.
However, randomized controlled trials are not always possible. Where the randomized allocation of support is infeasible, a quasi-experimental design may be appropriate. Like experimental designs, quasi-experimental evaluation designs compare outcomes of supported and unsupported firms but the allocation of support can be based on merit, as is conventionally the case, rather than randomly assigned. Quasi-experimental designs have their own requirements in terms of the identification of an instrumental variable that distinguishes between supported and unsupported firms, which may be difficult in some circumstances. But where such a variable can be identified, and suitable data is available, quasi-experimental methodologies are highly robust.
Please contact us if you would like to investigate experimental or quasi-experimental evaluation alternatives. Our network of leading researchers is ready to help you design your innovation support program to enable the best in the evaluation of program effectiveness.