Browse by author
Lookup NU author(s): Professor Habiba Al-ShaerORCiD
Full text for this publication is not currently held within this repository. Alternative links are provided below where available.
Winston Churchill once said that democracy is only as good as the education of the average voter. Equivalently, a meta-analyses that counts the votes or effect sizes of primary studies on a popular research question such as the relationship between the corporate social performance (CSP) and corporate financial performance (CFP) of may only be as good as the average empirical study. While many academics in the field of CSP reference meta-analyses as mighty evidence and Friede, Busch and Bassen (2015, JSFI) even conduct a meta-analysis of meta-analyses, Gray (2006, AAAJ) hinted towards a tautology with the same measure being used by some studies for CFP, while others use it for CSP. We statistically explore Gray’s tautology concern and also conduct further plausibility tests on all 21 meta-analyses published or circulating today. Our results do not find complete evidence in support of Gray’s tautology concern, but we identify more than 40 cases, in which meta-analyses coded the conclusion of the same study differently. In a stunning 17 cases, at least one meta-analysis coded a positive link between CSP and CFP, while at least one other meta-analysis coded a negative link. Hence, we conclude that the subjectivity included in the coding of most meta-analyses is significantly too large for such studies to be viewed as mighty evidence.
Author(s): Al-Shaer H, Hoepner A
Publication type: Article
Publication status: Unknown
Journal: 29th International Congress on Social and Environmental Accounting Research 2017, University of St Andrews, Scotland
Year: 2017
