You are here

  • > Publications
  • > Assessing the Validity of Evaluation Research by Means of Meta-Analysis

Assessing the Validity of Evaluation Research by Means of Meta-Analysis

Authors: Rune Elvik
Report nr: 430/1999
ISBN: 82-480-0091-5
Language: English
Attachments Summary
Full report
Sammendrag

This dissertation is based on seven appended papers, all published in scientific journals. The main research problem discussed in the dissertation is whether meta-analysis can be used to assess the methodological quality of evaluation studies. Illustrations of the use of meta-analysis for this purpose are given. The illustrations have been taken from road safety research. The dissertation discusses the concept of validity and proposes a set of formal criteria of validity for use in assessing the quality of evaluation studies. A distinction is made between four types of validity: Statistical conclusion validity, theoretical validity, internal validity and external validity. Twenty criteria of validity are proposed, of which nine criteria concern statistical conclusion validity, four refer to theoretical validity, four refer to internal validity and three refer to external validity. In the seven appended papers, these criteria are used systematically in meta-analyses of road safety evaluation studies. It is concluded that it is to a certain extent possible to assess the validity, and hence the methodological quality, of evaluation research within the framework of meta-analysis, but that one should not expect the use of meta-analysis to resolve all controversies surrounding this kind of research.

      

Gaustadalléen 21
0349 Oslo, Norway

Phone: +47 22 57 38 00
E-mail: toi@toi.no