Preparing Meta-Analysis of MetamodelUnderstandability

First Workshop on Empirical Studies of Model-Driven Engineering

Susanne Patig

Others  :  http://CEUR-WS.org/Vol-392/Paper2.pdf
PID  :  25797

来源: CEUR

【 摘 要 】

Metamodels are designed to be used by machines and humans. For human users, the understandability of the metamodel is important. Experimental investigations of understandability in computer science have led to conflicting results. To resolve such conflicts and gain insights into the nature of some phenomenon beyond singular experiments, meta-analysis can be applied, i.e., the statistical analysis of results obtained by other (primary) empirical studies. This paper shows the current obstacles for a meta-analysis of metamodel understandability: They consist in the heterogeneity of the individual experiments and deficient reporting. The paper provides a framework to increase the comparability of experiments on understandability. Such comparability enables future meta-analysis.