Data Reuse Experiences within Digital vs. Physical Zoological Collections

University of Michigan Museum of Zoology (UMMZ), February 20, 2014 Ixchel M. Faniel, Ph.D. OCLC Research fanieli@oclc.org Data Reuse Experiences within Digital vs. Physical Zoological Collections Elizabeth Yakel, Ph.D. University of Michigan yakel@umich.edu

For more information, please visit http://www.dipir.org Institute for Museum and Library Services (IMLS) funded project led by Drs. Ixchel Faniel (PI) & Elizabeth Yakel (co-PI) Studying the intersection between data reuse and digital preservation in three academic disciplines to identify how contextual information about the data that supports reuse can best be created and preserved. Focuses on research data produced and used by quantitative social scientists, archaeologists, and zoologists. The intended audiences of this project are researchers who use secondary data and the digital curators, digital repository managers, data center staff, and others who collect, manage, and store digital information.

Research Motivations & Questions What are the significant properties of quantitative social science, archaeological, and zoological data that facilitate reuse? 2. How can these significant properties be expressed as representation information to ensure the preservation of meaning and enable data reuse? Faniel & Yakel 2011

The Research Team

Research Methodology

Agenda • Snapshot of Users • Interviews • Observations • Discussion Image: DIPIR Team

A Snapshot Of 40 Data Reusers reuse data from online repositories and websites 90% 95% reuse data from museums and archives 65% are systematists 35% study ecological trends reuse data directly from colleagues 20% reuse data from journal articles 27.5%

The Discovery Process “I am a graduate student at [university], in Zoology and one of my committee members is an adjunct professor here, [name], so she noticed that I had genetic data for the same individuals that U of M has skull data for.” (CAU39) “… we started from that [author] paper and then added to it from other people’s work…So mostly from…reading other people’s papers.” (CAU22) “I knew from prior experience which museumshad large collections of material from the part of the world I was interested in.” (CAU19) “…that [aggregator repository]targets so many different collections that once you have access you know pretty much…You can identify very quickly what you need.” (CAU13)

Selection Criteria Condition of specimen Geographic precision Data coverage Physical variation of the species Matches another dataset Results of pre-analysis Availability of voucher specimen Location of repository Sequence has been published Manner in which the specimen is preserved Relevant taxonomically Availability of metadata Time period specimen collected

Interviews Image: DIPIR Team

Digital Data Selection Based On Locality …that’s the first filter…looking for specific species. And then for me, yeah, it’s been mostly about the geographic precision of the data, to say whether or not I can use that record for something. (CAU26). Image: Microsoft Clipart …often when it doesn’t meet my needs the most obvious reasons would be there’s just not enough data or it doesn’t cover…Like geographically it doesn’t cover the area I’m interested in well enough(CAU03).

Digital Data Selection Based On Other Datasets …we decide, okay, these Georeferences have an error that Is probably higher than, let’s say, five kilometers but our climate data is the resolution, the pixel size,…is may be 4.5 kilometers. So, anything that is above that size of pixel that we have, we actually cannot use. (CAU14) Image: Microsoft Clipart I include it [the sequence] in my dataset, do the analyses I’m going to do and then based on the results of those analysis look to see how those data match with the data that I’ve collected. (CAU05)

Trusting Digital Data “I can sort of qualitatively assess what the quality of taxonomic data might be just by it being, having some mention of the museum record. I know [a] …museum worker who is often... I don't know about an expert in say, my group, but at least has access to the relevant literature to make goodtaxonomic decisions about those fishes from which they took the tissue.” (CAU02) “I would go back to the literature to look at the paper it came from. I guess there is also to some degree the particular researchers’ that actually produced that sequence; I might actually know their reputations or what they kind of work on and trust it more or less.” (CAU12)

Trusting Digital Data “A lot of times, it's just a matter of looking at what the Latin name is that they supply because I can't really make a decision based on the information that I'm given. If I had a picture, I could use that when I'm taking into account their ability to identify something. But the main way that I do it is by looking at the geography of where they claim a specimen is located.” (CAU17) “Well, if there's a voucher specimenavailable then I can request that specimen from the museum where it's housed, re-examine it, confirm or deny that it is that particular species. If the voucher's there and it's the right species, then I have to go with it. If the voucher is not there, and I really question the identification…Because it's unreliable in my mind.” (CAU20)

Observations Image: DIPIR Team

Specimen Selection Based On Condition “It needs to be intact right? The skull needs to be intact. That isn't in the records usually, and I've gotten used to the idea that you just go and hope for the best, and figure that if they say they have 20, you might find six you could use. That would be a helpful thing to know.” (CAU34) Image: DIPIR Team

Specimen Selection Based On Holotypes For Comparison “[Many] holotypes from the past [are] deposited here in this collection. And then it's really useful to me, and important to make a comparison with those specimens that was the original description when the species already occur in the country. But to do that in the best comparisons, we need to compare morphological data with the new specimens that we already collected in the recent years.” (CAU29) Image: DIPIR Team

Deciding To Visit UMMZ “And it's a good-sized collection. Especially in terms of university's collections, there are a lot of specimens here, good taxonomic diversity, and it's also close for us . . . I'm going to the Smithsonian next week, but that's a lot more expensive, a lot more time consuming.” (CAU36) “I think it’s because I was a student here so I know, I knew what was here But I have to say, I worked on my dissertation in the same area, I worked on skull morphology, and so I learned as a graduate student that you go and find the museums that are most likely to have the specimens that you need.” (CAU34)

How Researchers Prepare For Their Visit To UMMZ “Well, the crucial thing there is getting a copy of the data associated with the specimens that are here…an Excel spreadsheet that gave all the information about the tissues that are held here and the morphological specimens. Using that database, I was able to then select which species we need to study.” (CAU32) Image: DIPIR Team

Interaction With Repository Staff “In this case, I was fortunate to have [UMMZ staff], who took the initiative to go through the collections and find the most well-preserved specimens that he could . . . So, actually looking through the collection that was done by [UMMZ staff] and he brought out the specimens for me to use. So, that aspect was alleviated by the fact that he gave me a lot of help.” (CAU33) Image: DIPIR Team

Discussion Image: DIPIR Team

Discussion • In global age of online databases people still need to see the actual specimens • Condition and depth of the collection is important • Aggregators vs. museum website vs. inventory system • Having data accessible online is great, but at times it just is not sufficient

Discussion • The discovery processes are similar but selection criteria are specific to research objectives • Gaining trust in data about the specimen from a distance

Acknowledgements Institute of Museum and Library Services Partners: Nancy McGovern, Ph.D. (MIT), Eric Kansa, Ph.D. (Open Context), William Fink, Ph.D. (University of Michigan Museum of Zoology) OCLC Fellow: Julianna Barrera-Gomez Doctoral Students: Rebecca Frank, Adam Kriesberg, Morgan Daniels, Ayoung Yoon Master’s Students: Jessica Schaengold, Gavin Strassel, Michele DeLia, Kathleen Fear, Mallory Hood, Annelise Doll, Monique Lowe Undergraduates: Molly Haig

Beth Yakel yakel@umich.edu Ixchel Faniel fanieli@oclc.org Questions? http://www.dipir.org

Data Reuse Experiences within Digital vs. Physical Zoological Collections