Methods & Statistics
Faculty of Social Sciences
On May, 25, 2018, Leonie van Grootel defended her thesis, titled:
Where No Reviewer Has Gone Before: Exploring the Potential of Mixed Studies Reviewing
The evidence-based movement has led to a large number of systematic reviews being produced (Dixon-Woods, et al., 2006; Petticrew & Roberts, 2006). Systematic reviews are used to determine effectiveness by aggregating the outcomes of evaluation studies, mainly randomized clinical trials (RCT’s). This approach has proven valuable in providing evidence for the question: ‘What works best to reduce problem X?’. Systematic reviews are characterised by explicit methods to the task, such as comprehensive searching, quality assessment of scientific studies and advanced analytical tools i.e. meta-analysis.
In policy-making and professional practice the need was felt to address other issues in addition to effectiveness, for example, how programs are received by target groups, how the program’s processes are linked to input and output, and what facilitates and obstructs implementation (Lomas, 2005; Dixon-Woods, et al., 2011). As a rule these questions match a qualitative methodology that is suited to describe and understand people’s experiences, considerations and decisions (Barbour, 2000; Harden et al., 2004). At the same time, qualitative research is often small-scaled and used to examine a specific, local context. However, when the available qualitative studies in a specific area are systematically synthesized, much more knowledge can be obtained than a single qualitative study can ever provide. The synthesis then covers larger and more diverse samples and more dimensions of the topic of interest (e.g. Van Wesel, Boeije, Alisic & Drost, in press).
By conducting a quantitative and a qualitative review on one topic, more and complementary knowledge can be gained when these reviews are integrated. This PhD-project focuses on the integration of quantitative and qualitative methods on the review level. Three methods that integrate evidence from qualitative and quantitative reviews are evaluated and further developed. The first method is based on the EPPI-approach, in which views of participants on the issue at hand are juxtaposed against effectiveness of an intervention. In the second method, the outcomes of the quantitative review will serve as a starting point of an exploration of the relations with the outcomes of the qualitative review. The third method consists of a Bayesian meta-analysis, in which we will use the outcomes of the qualitative review as starting point for the meta-analysis.
The project focuses on the development of synthesis methods, but the application of the project is on educational science. The topic of both reviews is collaborative learning in primary and secondary education.
Supervisors: Dr. H.R. Boeije, dr. F. van Wesel, prof. dr. J. Hox (Utrecht University)