Social scientists' data reuse behaviors: Exploring the roles of attitudinal beliefs, attitudes, norms, and data repositories

Abstract

A B S T R A C TMany disciplines within the social sciences have a dynamic culture of sharing and reusing data. Because social science data differ from data in the hard sciences, it is necessary to explicitly examine social science data reuse. This study explores the data reuse behaviors of social scientists in order to better understand both the factors that influence those social scientists' intentions to reuse data and the extent to which those factors influence actual data reuse. Using an integrated theoretical model developed from the theory of planned behavior (TPB) and the technology acceptance model (TAM), this study provides a broad explanation of the relationships among factors influencing social scientists' data reuse. A total of 292 survey responses were analyzed using structural equation modeling. Findings suggest that social scientists' data reuse intentions are directly influenced by the subjective norm of data reuse, attitudes toward data reuse, and perceived effort involved in data reuse. Attitude toward data reuse mediated social scientists' intentions to reuse data, leading to the indirect influence of the perceived usefulness and perceived concern of data reuse, as well as the indirect influence of the subjective norm of data reuse. Finally, the availability of a data repository indirectly influenced social scientists' intentions to reuse data by reducing the perceived effort involved.

What drives and inhibits researchers to share and use open research data? A systematic literature review to analyze factors influencing open research data adoption

“…Theory of Reasoned Action (TRA) Curty, Crowston [40] Theory of Planned Behavior (TPB) Harper and Kim [41], Joo, Kim [17], Kim and Adler [42], Kim and Yoon [43], Yoon and Kim [44], Zenk-Möltgen, Akdeniz [45] Technology Adoption Model (TAM) Yoon and Kim [44] Integrated Unified Theory of Acceptance and Use of Technology (UTAUT) with the two-stage expectation confirmation theory of Information Systems (IS) continuance Zuiderwijk [19], Zuiderwijk and Cligge [46] Institutional theory Kim and Adler [42], Kim and Yoon [43] Coordination Theory Zuiderwijk [19] Grounded Theory da Costa and Leite [47] Motivation theories (e.g. Expectancy Theory, Reinforcement Theory, The Multi-Motive Information Systems Continuance Model)…”

“…Research / academic discipline [17,40] -Disciplinary climate (a sense of community and openness to other researchers affiliated in the same field) [17] Research climate [43] Considered data reuse a prevalent research practice in their research communities [44] Traditions [40] Country [17] Sector [17] Requirements and formal obligations…”

“…Fun to explore data [18] Attitude (perceived concern) [17] Reinforces open scientific inquiry [50] Scientists' attitudes [40] Encourages diversity of analysis and opinion [50] Negative first impressions [58] Promotes new research [50] Stimulating economic growth, replication and validation of research [2] Enhancing transparency and reproducibility of the scientific enterprise [40] Scientist's beliefs and attitudes [40] Feeling worth (e.g., the feeling that the time spent on data reuse is time well spent) [40] Believe data reuse is good [44] Individual willingness [40] Replication of research results [11] Accelerate research [18] and increase the knowledge in the field [17] Advance our understanding of health and disease [18] Value users attach to being tested [18] Explore new interpretations of data [17] Intention to reuse data [17,43] Data being used enhances public trust and knowledge of the discipline [17] Provides a democratic scientific knowledge sharing platform: "Open access increases the pool of information available to anyone not just scientists" [4] Facilitating conditions Facilitating conditions [48] Lack of facilitating conditions [48] Digital tools (e.g. the possibility to involve more actors in data collection through citizen science platforms, not restricted by physical or cognitive distance) [4] Lack of availability of data [19] An open data infrastructure [19], a robust infrastructure for long-term use [50] Heavy reliance on the methods and techniques data producers employed to obtain, organize and code the data [40] The availability of data repositories [17,43,44] Technical bottlenecks…”

“…Human resource for question (advisors, data reuser groups, data producers) [43] Search options for open datasets are limited [19] Availability of internal resources [43] Searching for OGD in multiple languages is often not supported [19] Lack of support for data analysis [19] Interaction related to open data use is limited [19] Trust High level of trust [46] Trust concerns [19], perceived concern [43,44] Positive first impressions [58] Low level of trust [46] Improving data integrity [2,40] Unintentional misuse of the data [17,40,43] Data validity [58] Concerns about misinterpretation of the data [17] Trust in data producers [40] Open data can be reused for purposes they are not meant to be used for [19] Trust in the competence of the original investigator(s) (e.g. the original investigators' membership in a Community of Practice; appropriate educational training of the original investigator) [58] Transparent and honest attitudes of the original investigators [58] Reputation of the researchers who collected the data Trust in other researchers' measurement [17] Credible information availability [58] Good intentions and ethics of the original study that produced the data (e.g.…”

“…no commercial interests of the funder of the data; no apparent conflict of interest) [58] Study's funding sources [58] Existing evaluations of the data (e.g. many existing publications using the same data; large number of times the data has been reused and cited) [58] Expected performance Expected performance [48] Restrictions on use [50] Perceived (data) usefulness [17,43,44,46] Low perceived usefulness [46] Arrive to new findings [4], obtaining new insights [19,48] Potential waste of time [40] New scientific discoveries [2] Effort may be wasted on flawed data [40] Being aware of the state of the art and not reinventing the wheel [48] Negative reactions to data reuse [40] Feedback on the need for certain data and facilities [48] Issue of how to access usable citation and attribution information [55] Reproducibility of key research findings (and also experimental methods) that could push science ahead [4] Quality of reusing the data based on the context of the previous study [44] Allows collaboration across diverse groups [4] Inappropriate management or mistakes in management [58] Limited resources encourages collaboration [48] Original investigators' carelessness [58] Makes possible the testing of new or alternative hypotheses and methods of analysis [50], particularly when data are combined with other publicly available datasets [8] Risk of misinterpretation based on inappropriate use of data [59] Supports studies on data collection methods and measurement [50] Enables exploration of topics not envisioned by initial investigators [50] Permits the creation of new datasets when data from multiple sources are combined [50] Novel combinations of data …”