1 of 35

Racial Equity in Collections: Stepping Stones to Building Better Collections

Glenn Johnson-Grau, Loyola Marymount University, Head of Acquisitions and Collection Development

Amy Pham, SCELC, Database and eBooks Manager

Stepping Stones to Building Better Collections

Racial Equity in Collections:

2 of 35

There should be two piles of paper nearby with different colors.

Grab one of each color.

3 of 35

Sudden Selection!

You are a librarian with limited staff support and a scant supply of time.

4 of 35

Sudden Selection!

Your library has received special limited funding to add a new collection to support campus DEI initiatives.

You are a librarian with limited staff support and a scant supply of time.

5 of 35

Sudden Selection!

The funding has been allocated for a collection that is not in your area of expertise.

Your library has received special limited funding to add a new collection to support campus DEI initiatives.

You are a librarian with limited staff support and a scant supply of time.

6 of 35

Sudden Selection!

Collection A	Collection B
?	?
?	?
?	?

Based on the limited data provided, you have to choose between Collection A or Collection B.

7 of 35

Sudden Selection!

Collection A	Collection B
18,824 Titles	5,105 Titles
?	?
?	?

Based on the limited data provided, you have to choose between Collection A or Collection B.

8 of 35

Sudden Selection!

OAT = CHOICE Outstanding Academic Titles

Collection A	Collection B
18,824 Titles	5,105 Titles
322 Titles / 1.7% (OAT)	170 Titles / 3.3% (OAT)
?	?

Based on the limited data provided, you have to choose between Collection A or Collection B.

9 of 35

Sudden Selection!

Collection A	Collection B
18,824 Titles	5,105 Titles
322 Titles / 1.7% (OAT)	170 Titles / 3.3% (OAT)
Broad and comprehensive	Focused scholarly collection

Based on the limited data provided, you have to choose between Collection A or Collection B.

10 of 35

It's like comparing apples to dragonfruit.

11 of 35

The Big Reveal!

Collection A	Collection B
ProQuest Diversity, Equity and Inclusion Ebook Subscription	EBSCO eBooks Diversity & Ethnic Studies Subscription Collection

Collection A and Collection B are based on real collections:

12 of 35

What were we trying to find out?

More generally, could these collections present one stepping stone for diversifying collections?

Question 2

Could these collections help SCELC’s small academic libraries?

Question 1

13 of 35

Who are we?

Primarily private academics, <3000 FTE

330+ Institutions, 39 states

Opt-in licensing

Statewide California Electronic Library Consortium (SCELC)

Amy:

Short version:

My name is Amy Pham, and I’m the Database and eBooks Manager at SCELC, the Statewide California Electronic Library Consortium.

SCELC is a licensing-focused consortium that mostly serves small, primarily private academic, libraries. And when I say small, I’m not just referring to FTE.

�Long version:

So what is SCELC? SCELC is an opt-in consortium with a strong focus on licensing. LMU is a SCELC member library. Most of our member base is in California, but we represent institutions in 39 states, many of them private academics. 70% of SCELC’s represented libraries have an FTE of 3000 or less, many in that category with an FTE of less than 1200. And less than 3% of our schools are R1s. [So how does that translate? What does it actually look like to be a librarian at a SCELC library?] At an R1, you might have 86 full time library staff (including librarians). At a SCELC library, you would have 6. It’s the same work with less support.

�

14 of 35

Who are we?

At an R1 library:

86 FTE

At a SCELC Member library:

6 FTE

Statewide California Electronic Library Consortium (SCELC)

Median number of staff

15 of 35

Who are we?

Political Science,

African American Studies, Psychology

Glenn’s liaison areas

Assessing the Diversity of the E-Collection (ADER), 2017-2018

History and engagement with diversity in collections

SCELC member library

Loyola Marymount University (LMU)

Glenn talks here.

LMU is a mid-sized institution of about 9500 FTE, but in SCELC, LMU is at the bigger end

Range of reasons for LMU’s focus on diversity:

Our location in Los Angeles and the diversity of our student body and faculty.

LMU is a mission-driven institution. Both the Jesuit and Marymount traditions have focused on “social Justice” for a long time.

LMU library has focused on collection diversity for many years.

Assessing the Diversity of the E-Collection of the William H. Hannon Library (conducted in 2017-2018)

Results indicated that issues related to LGBTQ+ and disability were particularly underrepresented.

We made subsequent acquisitions with a special focus on those areas to enhance the collection.

Glenn’s liaison area

Political Science, African American Studies, Psychology

Previously Biology, Health & Human Sciences, Environmental Studies, Bioethics, etc.

I have been a “sudden selector” many times in my career.

16 of 35

Reflecting on shared motivations

SCELC

LMU

Collection evaluation

DEI Goals

Meeting user needs

Meeting member needs

Scalable solutions

Improving local collection

17 of 35

Reflecting on shared motivations

Systemic issues in libraries:

Systemic disadvantages for small libraries

Systemic issues in society:

Combating the pervasiveness of racism

Addressing

(Amy/Glenn/Amy)

�Amy: Coming from these mission-driven backgrounds at our organizations, Glenn and I were really motivated by shared perspectives on systemic issues that concern the library industry and that helped to inform the work we undertook for this presentation.

�Glenn:

The system disadvantages small libraries:

Issues for small libraries – the same as other libraries, but more so
The smaller (or more understaffed the library), the more it relies on outsourcing to vendors (or our users): approval plans, DDA/EBA, ebook packages
Like all libraries, small libraries strive to meet needs of users
A lot of the literature on diversity on collection development focuses on building and nurturing local expertise, community relationships, and other very labor-intensive activities
How can libraries do that with no money, no staff, and no time

�Amy:

Racism is a pervasive systemic problem that infects every single part of society, including libraries and the systems in which we operate. Whether we realize it or not, we have to be aware that racism has affected our collection development processes and policies, and has affected how we select books and which books get published. Glenn and I thought it was really important to directly address anti-racism as a uniting motivating factor in this presentation as a reminder of our shared long-term objective. Even though SCELC libraries might not have ample resources, they want to join all libraries in the fight against racism. So we really want to keep it at the forefront of our and everybody’s minds that any short-term solutions we propose as a result of the work we did for this presentation shouldn’t end up reinforcing existing racist practices in the long-term, especially if the short-term solutions end up not being so short-term.

��

18 of 35

Where to start?

Subscription collections as a bridge solution

Advantages of subscriptions

Low cost per title
Fill gaps in collections
Breadth/coverage outside core collecting areas
Subscription titles can be converted to owned titles
Dynamic!

Amy segues from Slide 13:

Where do we start addressing racism in our collections? I have to credit Nancy Kirkpatrick of OhioNet for this answer: Start anywhere.

�Glenn:

Does it make sense to use subscription collections as a bridge solution for addressing DEI in our libraries?
Building library collections is slow and labor-intensive
Advantages of subscriptions

Low cost per title
Fill gaps in collections by providing breadth/coverage outside core collecting areas
Subscription titles can be converted to owned titles where appropriate

This is important because subscription titles are not permanent

Vendors and publishers share this goal: They want us to convert subscriptions to purchases. There’s a reason they offer so many “convert to purchase” sales opportunities.
Dynamic! Collections can adapt to meet current trends and topics of interest.

19 of 35

Q1: Could these vendor products help SCELC’s small academic libraries diversify their collections?

Differences

Similarities

Subject coverage
Publication date range
Little overlap with flagship products
Little overlap with competitor products

Size
Content
Selection Methodology

20 of 35

Q1: Could these vendor products help SCELC’s small academic libraries diversify their collections?

Answer: Yes, with caveats.

21 of 35

Subject coverage

ProQuest

EBSCO

14 Subjects Total

8 Race & Ethnicity Subject Areas:

9 Race & Ethnicity Subject Areas:

18 Subjects Total

African American Studies

Asian American Studies

Asian Studies

Ethnic Studies

Global Black Studies

Hispanic & Latin American Studies

Jewish Studies

Native American Studies

African Studies

Asian & Asian American Studies

Black Studies

Indigenous & Native Studies

Islamic Studies

Jewish Studies

Latino and Hispanic Studies

Middle Eastern Studies

Amy:

�Along with the earlier stated advantages of subscription model ebook collections, these collections provide wide subject coverage for a multitude of topics related to diversity. Glenn and I chose to focus specifically on identifying titles related to race and ethnicity.

At the surface level, based upon the subjects covered in each collection, 50-60% of the titles cover different areas related to race. These are vendor-assigned subjects that Glenn and I categorized as Race & Ethnicity Subject Areas. ProQuest has 12 subjects (14 now), 8 of which are related to ethnic studies, and EBSCO has 18 subjects, 9 of which are related to ethnic studies. That’s significant representation for collections that are trying to be inclusive of the different facets of DEI.

You can see that based on the subjects listed, the coverage for each collection is pretty similar and hits on core curricular subject areas. Just based on this, the collections are reasonable short-term solutions for a small library that is just starting to assess the diversity of their collection. However, once we started digging into the title list, it became less clear whether the collections provided sufficient coverage of each subject area and whether the collections would sustainably augment a library’s existing collection. ��

22 of 35

Coverage gaps

The vendors were forthright about:

And yet…

Their goals for their collections

Selection criteria

Collection development model

Keywords for Disability Studies is in both collections
Keywords for Latina/o Studies is in neither
Keywords for Latina/o Studies is a Choice Outstanding Academic and RCL title

23 of 35

Overlap analysis:

Minimal overlap with flagship products or with each other’s products

No overlap with ProQuest Academic Complete or EBSCO eBook Academic Collection, by design.

Glenn

We did overlap analysis, and we will share the details with you in a handout in just a few minutes.

�Both vendors designed these collections to not overlap with their flagship products

From a business perspective, this makes sense

They don’t want to cannibalize sales and they want to protect customers from the ever-dreaded double-billing for the same books.

For small libraries, this is a serious problem:

They may not be able to afford flagship products

Librarians (and users) will not know if gaps developed or closely related content was excluded because it was in another of the vendor’s products, or for another reason, or no reason.

We would argue that for this type of collection that focuses on equity, this is a flaw: It prioritizes sales considerations over creating a coherent, clearly definable collection

24 of 35

Overlap analysis:

Minimal overlap with flagship products or with each other’s products

6% Overlap

Overlap 23%

77% Unique

94% Unique

ProQuest Diversity Collection

EBSCO Diversity Collection

Overlap with each other (1161 overlapping titles):

Amy

Not only was there minimal overlap with their own flagship products, there was also a low percentage of overlap with each other’s collections. ProQuest had a 6% overlap with EBSCO, EBSCO had a 23% overlap with ProQuest, although remember that EBSCO has a much smaller collection.

We were expecting an average closer to 60%, which is a rough estimate of the overlap between ProQuest Academic Complete and EBSCO eBook Academic. That means, even though their collections cover similar topics, they chose different titles to add to their collections.

Which honestly surprised us. And we’re not sure it was entirely by design. On one hand, this is a positive: the products are unique. If a library added both collections, they’d get 21,607 unique titles (assuming no overlap with their own collection). On the other hand, the lack of overlap is a negative: since these collections aren’t consistent with each other despite covering the same subject areas, the collections can feel like a random selection of “diverse” titles.

So to bring it back to our earlier point, “yes, but with caveats”.�

25 of 35

Q2: More generally, could these collections present one stepping stone for diversifying collections?

Answer: Not no.

26 of 35

Collection development methods

Standard approaches:

Assessing diversity in collections

Bibliographic lists

Award winners
Curricular title lists
Resources for College Libraries (RCL)

Holdings comparisons
Metadata-level assessment
Interdisciplinary LCC Assignments

Glenn:

There are standard methods for assessing library collections. The library literature indicates that these standard assessment approaches are adapted by libraries for assessing diversity in collections. This diversity assessment literature dates back more than 25 years, but has picked up steam in the last two years and include:

Bibliographic lists

Award winners
Curricular title lists
Resources for College Libraries (RCL)

Holdings comparisons

WorldShare Collection Evaluation

Metadata-level assessment

Titles, authors, publishers, place of publication, subject headings, LC classification

Interdisciplinary LCC Assignments

Alternative methods of classification to enhance access to interdisciplinary topics

�We found these standard methodologies of building collections were not sufficient/satisfactory when addressing DEI issues. This is not just true for the two collections under review. These are universal problems, so if both libraries and vendors are grappling this, there are opportunities for collaboration.

�Let’s look at the converse. Are there elements of the PQ & EBSCO methodologies that we found useful for additional methods for curating DEI-centric collection? (next slide)

27 of 35

Vendor selection processes

ProQuest

EBSCO

Selection Process:

Title assignment based on DEI taxonomy developed by internal group at ProQuest.
Ebook Central Advisory Group consulted.

Assigned subject specialist.
Title assignment based on GOBI metadata.
No specific advisory board consultation.

Amy:

For the Diversity, Equity and Inclusion Ebook Subscription Collection, ProQuest developed a proprietary taxonomy of DEI subjects and built it into the back-end of EBC [Ebook Central]. They used existing metadata in EBC to assign titles to each DEI subject in the taxonomy. The project managers in charge of the collection also regularly consulted with ProQuest’s Ebook Central Advisory Group.

In developing the eBooks Diversity & Ethnic Studies Subscription Collection, EBSCO assigned a subject specialist to the collection and focused on building the collection based on their familiarity with relevant publisher offerings. To make decisions about title assignments, they used metadata provided by the publisher, OCLC, and GOBI.

Though their selection processes are different, both ProQuest and EBSCO relied on similar metadata, with an emphasis on title and subject to make initial selections. No author identifying metadata.

How did EBSCO’s and ProQuest’s methodologies stack up to standard collection development methods?

Well, the Vendor selection process isn’t radically different from library collection development methods. They also used bibliographic lists and metadata-level assessment to make title selections.

But the vendor method is inverse to library method. The starting place is the same, but the end goals are different: vendors are trying to sell collections, and libraries are trying to acquire them.

Like the standard collection assessment methods, the vendor selection process wasn’t completely satisfactory either.

�

28 of 35

Collections that could have been

Examples

Collections that could have been:

Diverse publisher collection
Canonical historical collection

Collections have no clear identity

Vendors did not fully adopt standard collection development methodologies

Inconsistent application of stated scope

Glenn

Vendors used some but not all of standard collection development methodologies.

For example, they included OAT titles but inconsistently, as seen in the Keywords for series mentioned earlier

The result is that these collections have no clear identity: They are neither fish nor fowl

Inconsistent application of stated scope

�They could have chosen:

Previous ten years of highly diverse publishers, with a scholarly focus but adding publishers like in GOBI’s AP Plus! list (non-academic publishers of interest to academic libraries)
Or
Deeper historical collection with more canonical works

�Neither vendor wanted their collection to be pro forma. But that is how they felt to us and may to customers, particularly those concerned about the proliferation of vendor DEIA products.

A stronger curatorial hand would have made them better. Subject expertise is out there! Consult those subject experts!

29 of 35

Collaboration that could be

Assessing diversity in collections

Bibliographic lists
Holdings comparisons
Metadata-level assessment
Interdisciplinary LCC Assignments

Libraries can do more together

Collaboration required

Standard approaches:

30 of 35

Q2: More generally, could these collections present one stepping stone for diversifying collections?

Vendors have access to a lot of useful metadata
Inherent problems with market-based solutions

Answer: Not no.

(Amy)

Framing question: Is there anything in any of these methodologies that SCELC libraries or any small library could build on for future collection analysis?

�Like Glenn was saying, we’re not just here to criticize. We are also here to learn. There are things we can learn from the vendor selection process if we can address the inconsistencies.

�For one, vendors have access to a lot of useful metadata. They are able to harvest and classify metadata at a scale that a single library is not able to do efficiently. When Glenn and I tried to harvest work IDs for the overlap analysis, it took us 3 weeks and we ended up having to enlist the help of our local programmers.

If we were able to partner with vendors on metadata endeavors, it might be easier and faster to audit our collections. GOBI Interdisciplinary Tags are an existing example where metadata from a vendor is being used to help libraries identify diverse titles.

But right now, application of metadata, such as ISBNs, LC call number assignment, even publisher names! is inconsistent.

Secondly, there are inherent problems with market-based solutions. It can lead to:

-Reliance on generalized solutions that don’t fully serve the needs of your user community, problematic to tie social change to something with a financial bottom line

Diversifying collections is not the vendors’ responsibility. It’s our responsibility. But we can work together.

31 of 35

Our goal is long-term systemic change.

How can that change address the market?

32 of 35

How can that change address the market?

Starting with the ProQuest and EBSCO collections:

Consistent metadata
Transparency in bibliographic sources
Title lists by subject
Clarity about publisher relations

Partner with libraries

Clear way to provide feedback
Subject expert advisory boards

Amy:

�Starting with the ProQuest and EBSCO collections:

�Glenn:

Recommendations for vendors:

Use industry-standard, consistent metadata:

For example, OCLC Work IDs would allow libraries, and the vendors themselves, to compare editions to reduce duplication
We spent a massive amount of time trying to standardize metadata, which is a big disincentive to evaluating these collections, but also prevents library from clearly seeing their value

Amy:

Transparency in bibliographic sources

Using “award winners”? Which ones get selected?

Title lists by subject

Not just title list of whole collection but title lists with subject assignments indicated to make it easier to identify relevance to curricular needs

Glenn:

Clarity about publisher relations

There is no clarity in which books from which publishers get selected.

Amy:

Partner with libraries

-We’d like there to be a clear, formalized way to provide feedback. Right now, it’s on an individual library basis. We’d like to know that feedback is being received and the collections are responsive.

-And like Glenn was saying earlier, the expertise is out there. It’d be great subject expert advisory boards advising on the direction of these collections with the participants well-compensated.

[This slide is END of presentation. Next slide is “How could these products be better?”]

33 of 35

How could these products be better?

Open discussion: please provide feedback for vendors in the room!

Amy - (Keep an eye on the time - we want 10 minutes for Q&A)

Before we get to Q&A, we want to take a few minutes to ask you provide feedback for the vendors in the room. This is an open forum to talk about what you’d like to see improve with the EBSCO and ProQuest products we’ve discussed. But we don’t want to limit it. If you were involved in creating these collections from the ground up, what you have liked to see happen? What would your ideal collection have looked like?

�Jamie suggestion: giant sticky pads with this question written on it. Invite folks to leave suggestions afterwards.

�Subscription collection with EBA options to allow libraries to build local collections.

Clear direction of growth - knowing what we are investing in - what titles will be added in the future?

Better discovery

�If lull: For public domain canonical works, get quality editions! Not books from Open Road Media.

�Notes from SCELC folks:

Jason:

1) Be transparent about how the collection is built 2)Invite input on what could/should be added 3) Offer a competitive award for evaluation and criticism

(Suggestions for vendors)

�Donna LaFollette

I’d add Find consistency in your metadata - across vendors. It’s an endemic, persistent issue that causes libraries so many problems

��

34 of 35

Q&A + Closing Notes

Building expert lists, working on standardizing metadata, consortial collaborative collection development, and your ideas for how to assess diversity in collections

Efficiency reduces diversity. If we all buy the same “diverse” books, are we diversifying collections?

Future concerns:

Invitation to talk to us about:

35 of 35

Chat with us!

Amy Pham (SCELC), Database and eBooks Manager - amy@scelc.org

Glenn Johnson-Grau (Loyola Marymount University), Head of Acquisitions and Collection Development - gjohnson@lmu.edu

Available online here: https://bit.ly/3ymVBPp

Handout & Bibliography