Researcher Development Initiative (RDI), 2005 - 2010

A Multilevel Approach to Meta-Analysis for the Social Sciences

01/01/08 - 30/04/09


Meta-analysis is the most widely used method for quantitative systematic review in the Social and Health Sciences as a means of informing research, policy, and practice. However, comprehensive training in important new developments in meta-analytic methods (particularly multilevel and multivariate models with random effects) is not readily available in the UK.

This programme included a series of workshops on meta-analytic methods, from design through to interpretation of results, held in Oxford and Stirling. An emphasis was placed on producing a detailed set of train-the-trainer materials.


One day introductory seminars that aimed to develop general understanding of meta-analysis for researchers, lecturers, and policymakers who wish to engage with meta-analytic research or teach introductions to meta-analysis, but have limited/no understanding of the method.

Three-day workshops that aimed to cultivate skills in conducting meta-analysis for researchers and lecturers who have a sound knowledge of basic computational statistical methods, including experience with SPSS and an understanding of multilevel modelling (preferably using MLwiN).

A one-day advanced seminar, which looked at advanced issues in multilevel meta-analysis - delivered by a leading international expert - aimed at experienced meta-analysists who want to update their skills (preferably with some multilevel experience).


For further information on the course materials produced, please see


Principal Investigator: Professor Herb Marsh, University of Oxford
Co-Investigator: Dr Lars-Erik Malmberg, University of Oxford

Contact Details

Professor Herb Marsh, University of Oxford

Email: Herb.marsh@education.ox.
ac. uk
Tel: (0)1865 274041