More Data Please! Research Methods, Libraries, and Geospatial Data Catalogs: C-EBLIP Journal Club, August 25, 2016

by Kristin Bogdan
Engineering and GIS Librarian
University Library, University of Saskatchewan

Article: Kollen, C., Dietz, C., Suh, J., & Lee, A. (2013). Geospatial Data Catalogs: Approaches by Academic Libraries. Journal of Map & Geography Libraries, 9(3), 276-295.

I was excited to have the opportunity to kick-off the C-EBLIP Journal Club after a brief summer hiatus with a topic that is close to my heart – geospatial data! This article was great in the context of C-EBLIP Journal Club because it introduced the basics of geospatial data catalogs and the services around them, and provided an opportunity to look at the methods used by the authors as part of an ALA Map and Geospatial Information Round Table (MAGIRT) subcommittee research project.

Most of the group was unfamiliar with geospatial data catalogs, so the introductory material provided a good base for further discussion. There was good material about the breadth of the different metadata standards involved and how they are applied at the different levels of data detail. There was also good discussion about the importance of collaboration and the OpenGeoportal consortium in developing geospatial data catalogs.

One of the key themes of our discussion was that we would have liked to see more information about the research design and more data. We would have liked to see mention of the ethics process that the authors went through before carrying out their study. Our group had questions about the process that the subcommittee used to choose their sample, as it seemed like it was fairly limited. The authors acknowledge that this was not meant “to create a complete inventory” (p.281), but it seemed like it could have been broader to be more representative. We would also have liked to see the questions that were asked during the interviews and more of the qualitative data from the interviews themselves. It was unclear how structured the conversations with the catalog managers were and how the data presented in the tables and the conclusions were derived. The information presented in the tables was not consistently organized and seemed like it would have been more useful in the context of the interview. The pie chart they used on page 283 to show the “Approaches to Developing Geospatial Data Catalogs” was not as useful as a table of the same information would have been, as there are 5 pie sections to represent 11 data points.

In light of the questions around the data collection, the leap from the tables of responses to the recommendations seemed fairly large. In general, the lists of questions to consider when determining how to implement a geospatial data catalog were helpful but they aren’t really recommendations. The cases that they present provide some ideas about the staffing and skills required to create a geospatial catalog, but they are vague. The first case seemed unnecessary, as it states “The library has determined that there is a clear need to provide access to the library’s spatial data and other spatial data needed by the library’s customers. However, the library does not have the technology, staffing, or funding needed to develop a spatial data catalog.” It would have been nice to see some alternative solutions for those without the ability to create a full-blown data catalog like suggestions about some practices that could be put in place to start building the foundation of a geospatial data catalog like specific cataloging practices or file-type considerations.

Our discussion concluded with reflection on how carefully and critically we read articles in our general research lives. One of the great things about Journal Club is that we have the opportunity to really interrogate and dissect what we are reading. The ensuing discussion is an opportunity to see the article from many different perspectives. This makes us better researchers in two ways: we are trained to more thoroughly evaluate the things we read and we take that into consideration in the research that we produce.

This article gives the views of the author(s) and not necessarily the views of the Centre for Evidence Based Library and Information Practice or the University Library, University of Saskatchewan.

Walking the (Research Data Management) Talk

by Marjorie Mitchell
Librarian, Learning and Research Services
UBC Okanagan Library

Librarians helping researchers to create data management plans, developing usable file management systems (including file naming conventions), preparing the data for submission into repositories and working through the mysteries of subject-specific metadata schemes are at the forefront of the data sharing movement. All this work leads to research that is more reproducible, more rigorous, has fewer errors, and more frequently cited (Wicherts, 2011) than research that isn’t shared. In addition to those benefits, shared data leads to increased opportunities for collaboration and, potentially, economic benefits (Johnson, 2016). However, are we doing what we are asking our researchers to do and ultimately making our research data available and open for reanalysis and reuse? Are we walking the talk? Or is this the case of the carpenter’s house (unfinished) and the mechanic’s car (needing repair)?

When I’m speaking of data I use Eisner and Vasgird’s description of data as “a collection of facts, measurements or observations used to make inferences about the world we live in” (n.d.) because the research done by librarians consists of wide varieties of data: numerical, textual, photographic images, hand drawn maps, or diagrams created by study participants. Almost all have the potential to be shared openly and to act as a springboard for further research, subject to appropriate ethical considerations.

I started searching to see what data I could find from Canadian librarian researchers in repositories. I have not finished my search, but my early results show some interesting things. To date, this has not been a rigorous study, but more of a curious, pre-research “let’s see what’s out there” browse, and therefore must not be misconstrued as the basis for conclusions. I briefly looked internationally for a few studies and found a wider variety of topics with available datasets than I had found in Canadian repositories, which was what I expected to find.

Two things jumped out at me right away. First, when data is available, it is either from large, national or multi-institutional studies, or it is from studies that have been repeated over time, such as LibQUAL+®. Far fewer institution-specific or single researcher/research team datasets are “available.” Some of those have “request access” restrictions, meaning it may be possible to access the data with permission from the creator, but that is not guaranteed. The second thing I noticed was how difficult it is locate these datasets. Although there is a movement to assign unique and persistent identifiers to datasets, this has not, as yet, translated into a search engine that can comprehensively search for datasets.

I am happy to see a steady increase in the amount of librarian-generated research data being made available. Librarian-generated research is not alone in this trend. It is happening across the disciplines. While little library research is externally funded, it is worth noting some funders are requiring data management plans with the goal of data sharing. Some scholarly journals, particularly in the sciences, have strong policies about data sharing. Each change, minor or major, moves us more toward data that is shared as a matter of course, rather than data shared only reluctantly.

If this all sounds like “just another thing to do” or maybe “I don’t have the skills or interest to do this,” consider research data sharing as an opportunity to partner with another librarian who has those skills but perhaps lacks the research skills you have. Research partners and teams can allow people to contribute their best skills rather than struggling to compensate for their weaknesses throughout the process.

Finally, have a look at the data that is out there just waiting to be reused. Cite it, add to it (if allowed), and share your new results. I am confident this will add greater context to your research and highlight subtleties and nuances that might have remained invisible otherwise.

References

Eisner, R., & Vasgird, D. (n.d.) Foundation Text. In RCR Data Acquisition and Management. Retrieved from http://ori.hhs.gov/education/products/columbia_wbt/rcr_data/foundation/index.html

Johnson, B. (2016). Open Data: Delivering the Benefits. Presentation, London, UK.

Wicherts, J. M., Bakker, M., & Molenaar, D. (2011). Willingness to Share Research Data Is Related to the Strength of the Evidence and the Quality of Reporting of Statistical Results. PLoS ONE, 6(11). doi:hOp://dx.doi.org/10.1371/journal.pone.0026828

This article gives the views of the author(s) and not necessarily the views of the Centre for Evidence Based Library and Information Practice or the University Library, University of Saskatchewan.