News & Events

Department of Biostatistics responds to Statistics in Medicine tutorial

Updated 10:00 AM PST, Fri May 31, 2019


Below is the initial response from Statistics in Medicine to the department letter. We have, as they requested, altered the wording of our letter to reflect that the article was withdrawn rather than retracted. We await their full response, and hope that their internal discussions result in an apology

Response from Statistics in Medicine

We acknowledge your letter on behalf of the Department of Biostatistics at the University of Washington about the recently published and withdrawn tutorial: "How to investigate an accused serial sexual harasser." The Editors take the issues you raise very seriously. We are putting in place changes to some of our processes and will discuss the concerns raised in your letter. Once we have completed our discussions we will respond more fully.

We would also like to point out a distinction that you might not be aware of.  In your letter you state the article was retracted. In fact, the article was withdrawn. There are differences in how an article is treated in a publishing context.  In a withdrawal the published version of record is removed from the journal. This is not the case in a retraction.  We would be grateful if you could change retraction to withdrawal in your letter to fully reflect the process that has taken place.

We would also request that this interim response on behalf of the Editors be posted on your department website along with your letter. We would hope your department would agree that they would not want visitors to your website to think that we are ignoring the issues that you have raised in your letter.

We thank you for the interest you have shown in Statistics in Medicine and we would like to assure you that we are working hard to make the journal reflect the communities it serves.

Respectfully,
Ralph, Simon, Els, and Joel
Editors, Statistics in Medicine

Letter from UW Department of Biostatistics to Statistics in Medicine:

This letter has been modified as requested by Statistics in Medicine

To the editorial board of Statistics in Medicine:

On 21 March 2019, Statistics in Medicine published a Tutorial in Biostatistics titled How to investigate an accused serial sexual harasser [6]. This tutorial serves to undermine women. It is harmful to the statistical community, broader scientific community, and the general public. While the article was withdrawn recently, we believe that issues remain in its publication and the reasons given for its retraction. The University of Washington Department of Biostatistics requests that the editorial board issue a public apology and take steps to ensure that articles published in this journal reflect the journal’s goal to “influence practice in medicine and its associated sciences” [7] through accepted best practices in these fields.

We believe this article perpetuates the pervasive and unsupported myth that there is a large collection of innocent men targeted by an imagined “conspiracy of women” [3, 10]. In doing so, it implicitly denies the reality that in fact the opposite is true: the majority of incidents of sexual harassment and assault in the United States go unreported. A major reason for not reporting is fear of not being believed [9, 5, 2, 4]. The tutorial participates in this myth by mis-employing the statistical idea of equipoise: it ignores the totality of evidence, and proposes the test of an unrealistic null hypothesis of a conspiracy from women who lie about their experiences. Additionally, the authors use the phrase “self-perceived harassed women” in characterizing the alternative hypothesis; in effect, this phrase undermines the legitimacy of women’s narratives and experiences.

This tutorial is in direct conflict with the commitments to fighting sexual harassment made by many organizations, including the National Academies of Science, Engineering, and Medicine [8] and the American Statistical Association [1]. Additional concerns include: (i) the tutorial advertises commercial products STATXACT and Converus’ credibility assessment; (ii) the statistical methods involved are unnecessary and obfuscate that the tutorial is also an advertisement; and (iii) Converus’ credibility assessment is irrelevant to the stated focus of Statistics in Medicine.

The aforementioned problems with this tutorial raise serious questions about how Statistics in Medicine currently handles submissions. Although it has since been withdrawn, the reason given for withdrawal implies that the only issue with the tutorial was that it did not receive proper peer review. The tutorial’s publication still implies that the editorial staff deemed it appropriate for this journal and raises grave concerns with the current review process. We believe that these critical issues in the publication process must be addressed by the editorial board.

In light of these issues, we request that the editorial board issue a public apology to the readership (and broader community) for supporting the tutorial’s publication. We also request that transparent changes in the editorial and review process are enacted to ensure that all future published articles in Statistics in Medicine reflect the stated goals of the journal.

Signed,

The University of Washington Department of Biostatistics

*References

[1] American Statistical Association. ASA activities conduct policy. https://www.amstat. org/ASA/Meetings/Meeting-Conduct-Policy.aspx, 2018. [Online; accessed 24-April2019].

[2] L Langton, M Berzofsky, C Krebs, and H Planty, M Smiley-McDonald. Female victims of sexual violence, 1994–2010. US Department of Justice, Office of Justice Programs, Bureau of Justice Statistics       NCJ 240655, 2016. [Online at https://www.bjs.gov/content/pub/pdf/fvsv9410.pdf; accessed 26-April-2019].

[3] National Sexual Violence Resource Center. False reporting overview. https://www.nsvrc.org/publications/false-reporting-overview, 2012. [Online; accessed 26-April-2019].

[4] Rape, Abuse, and Incest National Network. The criminal justice system: Statistics. https://www.rainn.org/statistics/criminal-justice-system, 2019. [Online; accessed 26- April-2019].

[5] CM Rennison. Rape and sexual assault: Reporting to police and medical attention, 1992–2000. US Department of Justice, Office of Justice Programs, Bureau of Justice Statistics NCJ 194530, 2016. [Online at https://www.bjs.gov/content/pub/pdf/rsarp00.pdf; accessed 26-April-2019].

[6] JJ Shuster and M Handler. How to investigate an accused serial sexual harasser. Statistics in Medicine (withdrawn), 2019.

[7] Statistics in Medicine. Overview — aims and scope. https://onlinelibrary.wiley.com/page/journal/10970258/homepage/producti…, 2019. [Online; accessed 24-April-2019].

[8] The National Academies of Sciences, Engineering, and Medicine Action Collaborative on Preventing Sexual Harassment in Higher Education. Action collaborative on preventing sexual harassment in higher education. http://sites.nationalacademies.org/sites/ sexualharassmentcollaborative/index.htm, 2019. [Online; accessed 24-April-2019].

[9] The National Institute of Justice. Reporting of sexual violence incidents. https://www.nij.gov/topics/crime/rape-sexual-violence/Pages/rape-notifi…, 2010. [Online; accessed 26-April-2019].

[10] DA Weiser. Confronting myths about sexual assault: A feminist analysis of the false report literature. Family Relations, 66(1):46–60, 2017.

 

Questions: uwbiost@uw.edu