The perceived causes of research misconduct among faculty members in the natural, social, and applied sciences

ABSTRACT

Little research has investigated the conditions that lead to research misconduct. To develop effective intervention/prevention strategies, this void must be filled. This study administered a mixed-mode survey (i.e. mail and online) to a stratified random sample of tenured and tenure-track faculty in the natural, social, and applied sciences (N = 613) during the 2016–17 academic year. The sample includes scholars from 100 universities in the United States. Participants were asked about the extent to which they believe a variety of known criminogenic factors contribute to research misconduct in their field. Descriptive results show that professional strains and stressors (e.g. pressure to secure external funds and publish in top-tier journals) are most widely perceived to cause misconduct, followed by the low probability of detecting misbehavior. Results from the MANOVA model show that this pattern of perceived causes remains the same for scholars across scientific fields. Implications for future research are discussed.

KEYWORDS:

• Scientific misconduct
• academic integrity
• research fraud
• deviant behavior

Acknowledgement

The opinions, findings, and conclusions or recommendations expressed in this publication are those of the authors, and do not necessarily reflect those of the Department of Health and Human Services. The authors would like to thank Marcus Berzofsky, Katelyn Golladay, and Natasha Pusch for their valuable research assistance.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 Although scholars from other fields, such as mathematics, statistics, and logic, may also violate research ethics, these disciplines (which fall under the formal sciences heading) do not directly involve empirical inquiry. The current study focuses only on fields where the scientific method is used. Accordingly, law, literature, religion, music, art and many other disciplines widely considered part of the humanities are also excluded.

2 The administration of the email survey took place over four different time periods (i.e., 20 October 2016 [n = 67]; 17 November 2016 [n = 84]; 19 January 2017 [n = 89]; 1 March 2017 [n = 71]. At each period, 1000 sample members were emailed invitations to participate. A total of 139 of the 4000 emails (or 3.4%) reached the wrong person or were bounced back for technical reasons (e.g., closed account).

3 The mail survey was administered over two time periods (16 November 2016 [n = 147] and 17 January 2017 [n = 155]). Sixty of the 2000 surveys (or 3.0%) were returned because the member of the sample was not at the place of address or because the individual in question had moved without leaving a forwarding address.

4 Similar response pattern imputation, which is available in PRELIS 2.3 (Scientific Software International, Chicago, IL), was used to handle cases the missing values for the items in Table 1. This imputation method has been found to work well with survey data featuring ordinal response sets (Jönsson and Wohlin 2006). Prior to imputation, 1166 cells of the total 12,260 cells (or 9.5%) had missing values. Following imputation, 708 cells remained missing (or 5.8%).

5 It should be noted that the sample size increased slightly when estimating the confirmatory factor model in Figure 1. The increase is a result of the way Mplus handles cases with missing values (i.e., multiple imputation, MI). Technical details for this procedure are provided by Asparouhov and Muthén (2010).

6 It is important to note that administration of the online survey took place during and after the 2016 U.S. Presidential election—a period in which there was extensive media attention on the hacking of the Democratic National Committee’s computer system via a phishing email (i.e., clicking on a link in an unsolicited email; see Lipton, Sanger, and Shane 2016). While it is not possible to gauge the impact of such coverage on potential study participants, it is probably nontrivial.

Additional information

Funding

This work was supported by the United States Department of Health and Human Services, Office of Research Integrity (grant numbers ORIIR160028-04-00 and ORIIR150018-01).