Second retraction for former SUNY Upstate department chair found guilty of misconduct

Last week, we covered the case of Michael W. Miller, a former department chair at the State University of New York (SUNY) Upstate who was forced to retract a paper in the Journal of Neurochemistry after a university investigation found he had committed misconduct.

We figured more retractions might be on the way, so we weren’t surprised when a commenter informed us earlier today of “very interesting and odd retraction letter.” Miller has had at least one other retraction, it turns out, this one in Developmental Neuroscience for 2009’s “Lability of Neuronal Lineage Decisions Is Revealed by Acute Exposures to Ethanol.” Here’s the notice, published online on January 19:

On August 6, 2010, the Editor-in-Chief of Developmental Neuroscience , Steven W. Levison, was contacted by the legal counsel for SUNY Upstate Medical University and informed that an investigative committee at SUNY Upstate had determined that Dr. Michael Miller was guilty of multiple counts of research misconduct, and they requested that Developmental Neuroscience retract the 2009 article authored by Drs. Miller and Huaiyu Hu. The committee did not find Dr. Hu at fault for any scientific misconduct. After independently reviewing the reports of the investigative committee and the rebuttals provided by Dr. Miller, the Editor-in-Chief and the publisher concurred that there were errors in reporting sample sizes and in mathematical calculations that affected the statistical analyses. However, at that time it was our view that an erratum was more appropriate than a retraction of the article. Despite the flaws in the analyses of the data, it was our determination that those flaws could be corrected in a published erratum to produce a final product that would be more useful to the scientific community than a complete retraction of the work. On December 10, 2010, we requested that Dr. Miller provide a corrected figure with an explanation of the mistakes that had been previously published.

However, that erratum was never published as we were provided with additional information and testimonies that convinced us that the collected data may also be unreliable.

Accordingly, on October 18, 2011, we requested that a letter of retraction be provided that was co-authored and co-signed by both authors. It was our view that for an action as severe as an article retraction, both authors needed to sign the retraction letter. However, Drs. Miller and Hu were unable to agree upon a letter that could be jointly signed.

Therefore, at the request of the President of SUNY Upstate, with this statement we formally retract the article by Drs. Miller and Hu, Developmental Neuroscience 2009;31:50–57.

We have sent correspondence to both authors, to the President of SUNY Upstate and to Dr. Steven Goodman, Vice President for Research at SUNY Upstate, to inform them in advance that this action is being taken by the journal.

The paper has been cited just once, according to Thomson Scientific’s Web of Knowledge.

The level of detail in the notice is commendable, and as in the last retraction by Miller we covered, the journal takes pains to make it clear that his co-author is not responsible for the misconduct.

But what’s particularly interesting is that the journal chose to give Miller the benefit of the doubt — which we agree should be the default in most cases — even after being presented with what sounds like clear evidence of misconduct.  That’s noteworthy because journals often blame institutions for their delays in taking action, usually when those institutions drag their heels on investigations and won’t release the results when they’re complete. Here’s a case in which Upstate seemed to be transparent — to the journal, anyway; they haven’t made any public statements about Miller, as far as we know — and the journal needed to be prodded.

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.