Guidelines for Reporting Productivity Studies: A Review of the Reproducibility of DEA in the Service Sector

Becker Jörg, Beverungen Daniel, Breuker Dominic, Dietrich Hanns-Alexander, Rauer Hans Peter

Abstract
With respect to the scientific method, reproducing empirical research is a necessary step to develop generalizable knowledge. However, in service productivity measurement and managerial science in general, little attention seems to be paid on replicating empirical studies. One reason could be that many studies do not report their data and methodology in a degree of detail sufficient to reproduce them. In this paper, we provide evidence for a lack of reproducibility of empirical studies on service productivity measurement with Data Envelopment Analysis, based on reviewing service productivity measurement literature. The major impediments identified comprise a superficial description of the measurement process and data sources, insufficient details on the retrieval methodology, and an incomplete account of the extraction methodology. To improve the reproducibility of future studies, a structured set of guidelines is offered with which authors can report their studies more comprehensively.

Keywords
Data Envelopment Analysis; DEA; Service Science; Reproducibility