Scientists' perception on institutional data sharing support and pressure : Investigating ecologists’ data sharing behavior

Abstract: Data from underlying research has become increasingly important to scientists and the public in recent decades. As a result, funders and journal publishers have become increasingly demanding that scientists share their data. Universities have also been encouraged to advance their data sharing support units as a result of this development. Earlier studies on data sharing among scientists have primarily explored the barriers to data sharing, while motivations and perceptions among scientists that have shared their data have been examined less. To this end, this thesis investigates perceptions and responsibilities on data sharing among ecologists that have shared data in an open data repository within the last 12 months. As a public-funded and university-supported repository, the Swedish National Data Service (SND) is selected for this purpose. Semi-structured interviews with six ecologists are conducted to evaluate their motivations and perceived responsibilities on data sharing, The theory of planned behavior (TBP) serves as theoretical framework. Earlier TPB models are adapted to include new factors that potentially influence data sharing behavior among ecologists. The interviews highlight several individual and institutional factors that influence ecologists' data sharing in the SND repository. On the individual level, the informants perceive a strong personal responsibility to share their data publicly. On the institutional level, they perceive that journals have a large responsibility, and the findings indicate that journal pressure, in isolation, has a positive impact on data sharing. Perceptions about the SND support are ambiguous among the informants. While generally perceived as helpful and quality enhancing, most informants found it stressful and time consuming to share their data via the SND support. Thus, the combination of journal pressure and perceived stress associated with preparing data for SND has a negative influence on the motivation to share the data in the repository. Unless data sharing in SND is facilitated, or the pressure from journals is mitigated, scientists might increasingly opt for using repositories that require less strict metadata descriptions.