Abstract
Introduction
Data sharing has many benefits, including the ability to reproduce or verify research, the availability of publicly funded research to the public, which allows other researchers to ask new questions, and scientific advancement (Borgman, 2012). Despite these widely recognised benefits and pressure from research stakeholders to make data publicly available, there are still many factors that hinder data sharing at individual, institutional and international levels (Chawinga, 2019). In our study we seek to categorise the factors that influence data sharing, both positively and negatively, and to provide a framework of motives and barriers for further research.
Methods
Web of Science and Scopus were used to identify the primary studies. We performed the search using a broad search strategy, which yielded 1362 records. We also performed a Google Scholar search to identify studies not included in the previous searches. After deduplication and title screening for relevance, 146 papers remained, of which 83 studies were included in the analysis. The papers were analysed using a data extraction form and thematic coding. The codes were grouped into categories at a higher level of abstraction and finally a framework was constructed to show them in relation to each other.
Results
During the coding and analysis, a particular framework emerged, called HIFO after the four main groups of codes: Hopes, Incentives, Fears and Obstacles. Hopes are understood as positive outcomes that researchers hope to achieve by sharing data, either for themselves (citations and visibility), for the scientific community (transparency and integrity of research), or for the public. Incentives are such external stimuli that encourage researchers to share their data. Most of the categories of codes that fall into this group can work both ways, as both supporting and limiting factors. This is particularly the case for infrastructure, resources, regulations, formal recognition and the culture of data sharing. Fears can be defined as reactions to potential threats associated with data sharing, i.e. what researchers fear might happen to themselves or others (research participants) if their data are shared. Data misuse and misiterepretation are the most promint ones. Finally, barriers are understood as obstacles of an objective nature. They include lack of infrastructure, resources or recognition, perceived effort to collect and share data, and a weak culture of data sharing. Other barriers mentioned are lack of time, bureaucracy, the data itself, and power asymmetry.
Conclusion
Despite the undisputed benefits, data sharing is not as widespread as it could be for a variety of reasons. In our study, based on a qualitative review of primary studies, we sought to develop a framework that captures the incentives and barriers to data sharing in a holistic way.