Preparing Meta-Analysis of Metamodel Understandability

Author(s): Susanne Patig
Venue: Workshop on Empirical Studies of Model-Driven Engineering (ESMDE'08)
Date: 2008


DOI: (DOI link unavailable)
Type of Experiment: Experimental Research

This work presents new research in resolving conflicts among results of previously published empirical studies. The work focuses on the understandability of metamodels through use of statistical analysis of past results, called meta-analysis. The basic steps of which meta-analysis is comprised are listed along with the requirements which the included studies must satisfy.

From their analysis, the authors present "current obstacles for a meta-analysis of metamodel understandability," namely "heterogeneity of the individual experiments and deficient reporting." To combat these obstacles, a taxonomy is given in the form of a framework which helps generate lists of variables and control techniques. These lists can then be used in "methodologically equalizing differences among experiments," allowing the comparison of those experiments.