Data Management and Data Sharing in Developmental Science

• In Event: Are We Encouraging Open, Reproducible, and Inclusive Science? The Metascience of Developmental Science

Abstract

Background: Data sharing has been a requirement of many projects funded by the Institute of Education Science and the National Institutes of Health for several years. However, a recent survey of education researchers by Makel et al., (2021) found that 80% of responders reported never or occasionally sharing their data. At the same time, 75% of researchers reported that data sharing should be practiced always or often. This disconnect suggests that there is a general lack of understanding around data sharing, and that education researchers have some salient barriers to adopting this practice.
One potential barrier is the lack of expertise in the management of their data; research teams do not have sufficient documentation or data that they believe is clean enough to share. To that end, our study sought to better understand the data management and data sharing practices of child development researchers. We will present the findings on types of data management activities researchers perform, their confidence in those activities, and how they have been trained in data management. Further, we examine the prevalence of both positive attitudes and perceived barriers to data sharing practices. We also seek to understand the relations among these two perceptions; whether those who feel less confident in data management identify different barriers to data sharing.
The main research questions that will be assessed in this study are: RQ1: How do education researchers engage in data management practices? RQ2: What barriers do education researchers perceive as being salient to their work? RQ3: Are specific barriers more salient for researchers who do not see the benefits of data sharing?

Method: Sample. This study targeted researchers working in the field of education science. To be eligible to participate, we required researchers to self-identify as a scientist working in the field of education who had served as the primary investigator of at least one study. We used email, social media, and listservs to contact potential participants for the study. Measures. We developed a series of items associated with barriers towards data sharing. Participants were asked to what degree they agree or disagree positive items such as “data sharing is good for my career”, as well as several potential barriers such as “my IRB will have a problem with me sharing data”. Analysis. RQ1 and 2 were analyzed using descriptive statistics, RQ3 was addressed with a regression model.

Results: Data from 202 researchers were included. We found that data management is most often performed by researchers themselves or their graduate students, and the majority of respondents (54%) reported that they had not yet shared data. The most salient barrier to data sharing was the belief that others might misinterpret their data (60% agreed).

Conclusion: Our work suggests that there are still many salient barriers to researchers adopting the practice. However, many of the negative attitudes and beliefs expressed in items within the survey are barriers to open science practices that can in fact be overcome, and resources exist to aid in that education.

Authors

• Jessica Logan, Vanderbilt University
  Presenting Author

• Busra Ceviren

• Ally Hanson

• Lexi Swanz