‘The methodology does not generate the results’: Journal corrects accounting study with flawed methods

What a difference a Yi,t=β0+β1IOˆi,t+β2Xi,t+ωt+εi,t.Yi,t=β0+β1IO^i,t+β2Xi,t+ωt+εi,t. makes.

The authors of a 2016 paper on institutional investing have corrected their article — to include the equation above — in the wake of persistent questions about their methodology. The move follows the protracted retraction earlier this year of a similar article in The Accounting Review by the duo, Andrew Bird and Stephen Karolyi, of Carnegie Mellon University in Pittsburgh, for related problems.

The bottom line, it seems, is that Bird and Karolyi appear to be unable adequately to explain their research methods in ways that stand up to scrutiny. 

The correction involves a paper published in The Review of Financial Studies, from Oxford University Press, titled “Do institutional investors demand public disclosure. According to the statement (the meat of which is behind a paywall): 

A special committee appointed by the SFS investigated the paper “Do Institutional Investors Demand Public Disclosure,” by A. Bird and S. Karolyi, published in the Review of Financial Studies in 2016 (10.1093/rfs/hhw062, volume 29, 3245–3277), after receiving reports on the inconsistency in the paper’s use of Russell vs. CRSP market capitalization ranks between the published version of the paper and its previously publicly distributed working paper version. Using the data and code provided by the authors, the committee determined that the estimates based on the use of Russell vs. CRSP ranks were not substantially different from each other. However, the Committee discovered that equations (1) and (2) on page 3254, which describe the two-stage model, do not describe the actual regressions that were used to generate the results presented in the paper. Moreover, the committee found that the paper’s main inferences are not robust if the models are estimated as described on page 3254 of the paper. In summary, the Committee concluded that the methodology described in the paper does not generate the results reported in that paper. Futhermore, the Committee concluded that the actual specification the authors acknowledged to have used in the paper was econometrically inconsistent. This is because the missing term constitutes an “excluded variable” which was not justified by the setting or discussed in the paper. The authors’ misstatement prevented the issue from being discovered in the review process.

We emailed Karolyi and Bird for comment but have yet to hear back. We also reached out to Itay Goldstein, the executive editor of the journal, to find out why it opted for a correction rather than a retraction. 

Alex Young, an accounting researcher at Hofstra University in Hempstead, NY, who raised questions about Karolyi and Bird’s retracted article and ultimately failed to replicate it, was not one of the readers who raised concerns about the other article. But, he told us: 

I would be very interested to see the authors’ data and code that  generate the results presented in the paper.

Like Retraction Watch? You can make a tax-deductible contribution to support our work, follow us on Twitter, like us on Facebook, add us to your RSS reader, or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at team@retractionwatch.com.

One thought on “‘The methodology does not generate the results’: Journal corrects accounting study with flawed methods”

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.