1 METAEVALUATION AS A TOOL FOR THE IMPROVEMENT AND DEVELOPMENT OF THE EVALUATION FUNCTION IN PUBLIC ADMINISTRATIONS María Bustelo (Associate Professor. Universidad Complutense de Madrid) Departamento de Ciencia Política y de la Administración II Facultad de Ciencias Políticas y Sociología Universidad Complutense Campus de Somosaguas s/n. 28023 MADRID mbustelo@cps.ucm.es Nowadays, in most European countries, evaluations are becoming a reality, so that there is a certain increase on empirical references of evaluation experiences in different policy fields. However, most of these experiences have not been already studied in a systematic way. In other words, there are very few investigations which have the evaluations that are being done in a concrete field as their main object of study. Doing metaevaluations can be an original approach for these kinds of studies. A metaevaluation is defined here as the systematic gathering, analysis and assessment of a pre-determined set of evaluation processes. Thus, the (meta)evaluation can be a useful tool for the best understanding and improvement of the evaluation function around a concrete policy sector. An in-depth review of the concept, aims and functions of metaevaluations is done. It is distinguished from other related concepts, such as evaluation synthesis and meta-analysis. We also identify different types of metaevaluations, define (meta) evaluation criteria for the assessment of evaluation processes, and elaborate on the strenghts and weaknesses of these kinds of studies. Finally, this paper aims to discuss and reflect on the lessons learned and method of metaevaluation applied in a metaevaluation on gender equality policies in Spain, conducted in 2000-2001. Meta-evaluation and evaluation synthesis: concept and differences The concept of meta-evaluation was first developed by Michael Scriven, as early as 1969, although an articulated definition by this same author can be found later on in his “Evaluation Thesaurus”. For him, meta-evaluation is: “the evaluation of evaluations - indirectly, the evaluation of evaluators - and represents an ethical as well as a scientific obligation when the welfare of others is involved. It can and should be done in the first place by an evaluator on his or her own work; although the credibility of this is poor, the results are considerable gains in validity...[Because] the results of self-evaluation are notoriously