Methodological rigor of corporate governance studies: A review and recommendations for future studies

Abstract

Abstract Manuscript Type Article/Review Research Question/Issue This study presents the results of a longitudinal content analysis of governance articles published in both CGIR and other leading management outlets. Our aim with this review is to identify problematic areas, as well as opportunities to improve the methodological rigor of governance studies. Research Findings/Insights We examine key design aspects of corporate governance articles. We identify and compare trends (1) over time and (2) across publishing outlets. Specifically examined are the geographic scope of samples, design characteristics (i.e., single‐ versus multi‐year data, the use of archival, survey, experimental, and qualitative designs, data sources, and reliance on concurrent versus longitudinal data), survey response rates and the use of tests for non‐response bias, the statistical analysis techniques used, reporting of results (number and proportion of hypotheses supported, non‐significant, and counter‐significant, use of statistical power analysis, reporting of descriptive statistics and correlations, use and significance of control variables, and endogeneity controls), and the sophistication of the models tests (i.e., mediation and moderation). Theoretical/Academic Implications Based on the results of the review, recommendations for academic researchers are offered in the following areas: (1) designing theoretical tests which both incorporate and go beyond main effects; (2) evaluating a research design's analytical parameters on an a priori basis to maximize reliability, validity, and the potential for publication; (3) incorporating assessment of sensitivity and robustness checks into the analysis; and (4) comprehensive reporting of results including post hoc power analysis, complete and expanded correlation matrix(es), reporting and assessing counter‐results, and avoiding HARKing.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 33 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.