Data withholding in academic genetics - Evidence from a national survey

Context The free and open sharing of information, data, and materials regarding published research is vital to the replication of published results, the efficient advancement of science, and the education of students. Yet in daily practice, the ideal of free sharing is often breached. Objective To understand the nature, extent, and consequences of data withholding in academic genetics. Design, Setting, and Participants Mailed survey (March-July 2000) of geneticists and other life scientists In the 100 US universities that received the most funding from the National Institutes of Health in 1998. Of a potential 3000 respondents, 2893 were eligible and 1849 responded, yielding an overall response rate of 64%. We analyzed a subsample of 1240 self-identified geneticists and made a limited number of comparisons with 600 self-identified nongeneticists. Main Outcome Measures Percentage of faculty who made requests for data that were denied; percentage of respondents who denied requests; influences on and consequences of withholding data; and changes over time in perceived willingness to share data. Results Forty-seven percent of geneticists who asked other faculty for additional information, data, or materials regarding published research reported that at least I of their requests had been denied In the preceding 3 years. Ten percent of all post-publication requests for additional information were denied. Because they were denied access to data, 28% of geneticists reported that they had been unable to confirm published research. Twelve percent said that in the previous 3 years, they had denied another academician's request for data concerning published results, Among geneticists who said they had Intentionally withheld data regarding their published work, 80% reported that it required too much effort to produce the materials or information; 64%, that they were protecting the ability of a graduate student, post-doctoral fellow, or junior faculty member to publish; and 53%, that they were protecting their own ability to publish. Thirty-five percent of geneticists said that sharing had decreased during the last decade; 14%, that sharing had increased. Geneticists were as likely as other life scientists to deny others' requests (odds ratio [OR], 1.39; 95% confidence interval [CI], 0.81-2.40) and to have their own requests denied (OR, 0.97; 95% CI, 0.69-1.40). However, other life scientists were less likely to report that withholding had a negative impact on their own research as well as their field of research. Conclusions Data withholding occurs in academic genetics and it affects essential scientific activities such as the ability to confirm published results. Lack of resources and issues of scientific priority may play an important role in scientists' decisions to withhold data, materials, and information from other academic geneticists.