Citation

Sampson A. Evaluation (Sage) 2007; 13(4): 477-493.

Copyright

(Copyright © 2007, Tavistock Institute of Human Relations, Publisher SAGE Publishing)

DOI

10.1177/1356389007082132

PMID

unavailable

Abstract

Using 'evidence' to falsify rather than verify patterns in data and searching for alternative explanations enables a better understanding of the circumstances that explain why and how a social programme works or does not work. An analysis of the extent to which a programme is meeting its aims and objectives to find out if it provides a solution to the policy problem, is more rigorous.The roles researchers adopt influence the quality of an evaluation; facilitating a better understanding of the theories embodied in programmes enhances an evaluation while being a 'broker of compromise' can limit access to information. Researchers have a valuable role in promoting learning. A robust evaluation framework integrates strategies for generalizing at the outset and identifying mechanisms of change or causal mechanism is a way forward. Examples are taken from recent evaluations conducted by the author and colleagues to illustrate the arguments.