Reuse of research data is an important topic across disciplines, where repurposing data can lead to new discoveries. Open science initiatives and tools encourage data sharing for reproducibility and reuse, a practice that has been adopted to varying extents throughout ecological research communities. A recent survey of astronomers conducted by the author of this poster found that 34.6% of peer-reviewed journal articles included in the study (n=211) were produced using data discovered serendipitously or otherwise through review of published literature. These data are either available in openly accessible locations or obtained by asking authors directly for their data. Other researchers have also found that the literature is a frequent gateway to research data for secondary analysis, including for ecology researchers. To broadly assess the accessibility of data behind papers, and to explore how researchers across disciplines recognize, assess and locate potentially-useful data while reading the literature, a study is underway to compare the information behavior of astronomers and ecologists. “Think-aloud” interviews have already been conducted with a convenience sample of 7 astronomers as they read a total of 32 journal articles and attempted to locate the underlying data online. A sample of 7-10 ecologists will be recruited for additional interviews.
Results/Conclusions
Interviews with astronomers were analyzed using protocol analysis techniques. Think-aloud statements were coded according to a pre-existing taxonomy of information-seeking strategies that was adapted for this study, including the following search facets:
- Scan: Scanning the text of the article
- Search: Using online resources to search for relevant information
- Learn: Developing an understanding of the underlying data while reviewing the article
- Select: Accessing relevant information about the data
- Recognize: Recognizing characteristics of the research and data
- Specify: Specifying datasets, archives, instruments and software for retrieving data
- Information: Retrieving underlying data
- Meta-information: Identifying characteristics and/or assessing quality of data
As a preliminary analysis showing how astronomers moved from one search strategy to another, the transitional probability was calculated and a network analysis was conducted. Results show that interviewees frequently moved from one facet back to the same facet, as with Search (P=.3874). Movement to Select was common, while movement to Information and Metainformation (categorized as actually finding and assessing the located data) was uncommon. Specify was found to be an important intermediary step between facets. This study overall aims to inform domain-spanning infrastructures and data discovery interfaces, to effectively “harness the ecological data revolution”.