Negotiating the Intersection of Science and Commerce

In the accompanying article, Abel, Jordan, Harlow, and Hsu (2018) describe the results of a new study examining the utility of a screening measure design to detect risk of sexual offending for job applicants interested in working with children. Using data from a large sample of cases that completed the proprietary measure, an algorithm was developed that, after setting high specificity values, identified 77% of men and 72% of women applicants who were considered to pose a risk of child sexual exploitation or child abuse in the workplace because their responses were similar to individuals who had sexually offending against children and different from those with no evidence of sexually offending against children.

This study clearly addresses an important topic regarding child protection and prevention of sexual exploitation and abuse, especially in light of ongoing revelations regarding child sexual abuse occurring within institutions (indeed, an upcoming special issue of this journal). Moreover, independent reviewers thought the study was suitable for publication in Sexual Abuse, and I agreed.

When first submitted, I was concerned that the review and editorial process could not be fully executed given the proprietary nature of this measure, which resulted in the manuscript not providing all the details one might expect in a scientific study. As Abel et al. note in their paper, there was also concern that any value of the screening measure would be lost if details about the items and the scoring were to become publicly known. As Editor-in-Chief, I sought a solution that would allow the proprietary nature of the items and scoring to be protected, while still allowing for independent review, verification, and replication.

A process was developed in consultation with the authors and the associate editors of this journal, both for the review and for readers if the manuscript were to be accepted for publication. This process involved providing the data and syntax, in confidence, for the associate editor and reviewers to be able to evaluate this manuscript and understand the results. In addition, the authors agreed to explicitly recognize the importance of postpublication verification by including an author note explaining the data and syntax could be reviewed by readers upon the signing of a nondisclosure agreement: This agreement prevented the receiver from sharing the data or revealing the items or scoring of items, but not preventing the receiver from discussing the data or any criticisms.

This process was developed for this particular submission, but will be used for future submissions involving proprietary content, when applicable. My hope is that this is a workable model for how to negotiate the tensions between scientific and commercial interests. This is not the first time this type of tension between science and commerce has come up, and other journals have addressed it in different ways, including disclosure of conflicts of interest (which we have) as well as expected reporting guidelines (which we added in June 2017; see Seto, 2017). In addition to consultation with the associate editors and peers, I also considered guidelines and recommendations, including those of Kanter (1998), the American Psychological Association (https://www.apa.org/science/leadership/bsa/data-sharing-report.aspx), PLOS (http://journals.plos.org/plosone/s/materials-and-software-sharing), and the conflicting interests policy of this journal’s publisher (https://us.sagepub.com/en-us/nam/declaration-of-conflicting-interests-policy). I welcome reader comments and suggestions regarding this model, as well as other journal guidelines or policies.