Introduction

Open Educational Resources (OER) are defined by UNESCO as “teaching, learning or research materials that are in the public domain or released with intellectual property licenses that facilitate the free use, adaptation and distribution of resources” (UNESCO 2021). The use of OER textbooks and course materials in higher education is associated with improved grades and retention, particularly for low-income students, part-time students, and students who are historically underserved in higher education (Colvard, Watson, and Park 2018; Fischer et al. 2015). Additionally, OER have been shown to positively impact teaching and learning, supporting more inclusive, collaborative, and student-centered pedagogy (Griffiths et al. 2022; Petrides et al. 2011). As such, increasing the use of OER in college courses has the potential to broadly and positively support equity in higher education. However, an ongoing challenge and often primary reason for higher-education faculty reluctance to adopt OER in their courses is the difficulty in discovering and identifying high-quality OER textbooks and materials (Abeywardena and Chan 2013; Belikov et al. 2016; Judith and Bull 2016; Lou et al. 2020; Perifanou and Economides 2022).

Academic librarians worldwide have long had a role in sourcing high-quality open course materials for both students and faculty members (Martin 2010; Bell 2015; Smith and Lee 2016, mwinyimbegu 2018). Librarians also play a critical role in OER adoption and sustainability, through engaging in advocacy and education; providing strategies to faculty to find and evaluate OER; maintaining OER via institutional repositories and collections; using metadata, indexing, and classification skills to improve OER access; and leveraging copyright and media expertise to facilitate the curation and creation of OER (Okamoto 2013; Smith and Lee 2016; Reed 2019; VanScoy 2019). However, librarians also struggle with the OER discovery process, which often involves searching multiple repositories and is hindered by insufficient or inconsistent metadata (Sobotka, Wheeler, and White 2019). The documented struggle by both faculty and librarians in their OER searches suggests the need for significant back-end and front-end design improvement in OER repositories and online libraries to improve the discovery and curation process. One way to support this design process is through the development of data-driven faculty and librarian user personas, projections of different kinds of potential users that highlight specific goals, needs, and challenges of faculty and librarian OER seekers.

The Institute for the Study of Knowledge Management in Education (ISKME) is a non-profit organization that hosts the free, public digital OER library, OER Commons. ISKME also creates and hosts digital library microsites (unique, indexed libraries of OER that are locally administered) for partnering institutions, frequently academic library consortia. As part of a larger Institute of Museum and Library Sciences (IMLS) grant–funded effort to design sustainable OER discovery solutions that build toward a national digital infrastructure for OER, ISKME implemented a user research study to identify the specific OER search and metadata needs of postsecondary faculty and librarians in collaboration with six partner library consortia: the Louisiana Library Network (LOUIS), the Ohio Library and Information Network (OhioLink), Virginia’s Academic Library Consortium (VIVA), the Private Academic Library Network of Indiana (PALNI), the Pennsylvania Academic Library Collaboration, Inc. (PALCI), and the Digital Higher Education Consortium of Texas (DigiTex). More specifically, this study aimed to answer the following research questions in order to inform OER discovery design:

• What are the tasks and decision-making processes faculty and library staff use when selecting, evaluating, and assembling both individual OER and collections of OER?
• What augmentations to existing metadata are needed to accommodate their decision making?
• What pain points do they encounter in the OER curation process?

Findings from in-depth interviews with faculty and librarians with moderate to high OER experience were then used to develop curation personas to guide the formulation of design solutions for OER metadata and discovery for higher-education faculty and librarians.

OER discovery

While researchers have noted that current industry standards for content search and discovery do not provide satisfactory mechanisms for identifying high-quality learning objects suitable for specific teaching contexts (Limongelli et al. 2022), and acknowledge the importance of incorporating task-based user-experience research in to the development of future solutions (Cortinovis et al. 2019), there has been little published research that delves into the specific information needs of faculty and librarians in their OER search and curation processes. Speaking to their own experiences, Oregon-based academic librarians Sobotka, Wheeler, and White note, “While there is ongoing improvement in some of the larger open educational resources (OER) search engines, librarians sending emails to listservs asking ‘anyone know of OER on this topic?’ and keeping old-fashioned reading lists of valuable OER are common occurrences.” They attribute this reliance on search methods outside of OER repositories and libraries to “subpar and variable” metadata. In their case study of searching for language learning OER, Perifanou and Economides (2022) point to lack of uniformity in repositories and in metadata, inaccurate and obsolete metadata, poorly or inaccurately labeled copyright information, and outdated OER with broken links as contributing to OER discovery barriers.

OER perception studies give hints as to what elements of discovery and evaluation are challenging to faculty, and what metadata they might find useful in their search and curation process. More than one study suggested that an underlying challenge may be that many faculty are not aware of existing OER repositories (Abeywardena, Gajaraj, and Chan 2012; Belikov and Bodily 2016). Lack of repository awareness however, is presumably less likely to be true of librarians. OER perception studies also suggest that even when faculty and librarians do know where to search for OER, they may not find the information needed to adequately assess whether the OER they have found is suited to their purpose. Abeywardena, Gajaraj, and Chan’s 2012 survey of academics in five countries in Asia indicated that discovery concerns included the need to find OER that is “specific, relevant, and quality” (6). In a more recent US-based multi-campus survey of faculty perceptions of OER and impediments to use, Elder et al. (2020, 141) found that 86% of faculty listed one or more of the following challenges: “finding comprehensive materials, finding suitable materials, finding high-quality materials, finding up-to-date materials, and finding locally relevant materials.”

Existing information on discovery barriers that have been encountered during specific search attempts is frequently in the form of internal feedback mechanisms rather than published literature. For example, ISKME and its microsite partners conducted internal surveys on faculty and library discovery needs. These surveys indicated that:

• the curated collections often do not enable faculty users to efficiently identify how well OER aligns to specific course requirements or to understand what adaptations are needed to meet those requirements;
• faculty end users lack social endorsement of material quality by way of peer feedback and reviews of the resources;
• while full, textbook-level OER resources are now more discoverable for faculty due to curation and alignment efforts, identifying and sharing OER resources in smaller content chunks (such as lectures, slides, videos, and assessments) remains challenging due to lack of standards and coordination in metadata tagging strategies;
• the process of finding, aligning, and maintaining OER for their centralized OER repositories is time consuming, difficult, and unsustainable by current library staff; and
• overall, the consortia are not able to grow their own collections by leveraging the work of their peers because their process lacks an efficient way to ingest OER being curated by each other and by external OER repositories.

Past and present attempts to remediate discovery challenges have often focused on the creation of cross-repository search engines like the Learning Registry (Cavanaugh 2018), Mason Metafinder, or MERLOT’s Smart Search; methods to improve automated processes for harvesting and sharing metadata and assessing OER quality (Koutsomitropoulos 2010; Piedra et al. 2014); or ways to improve and standardize OER metadata within and across repositories (Bothmann 2020; Gallant et al. 2022). One well-known example of an approach to standardizing metadata is the work of the OER Discovery Working Group, sponsored by the Scholarly Publishing and Academic Resources Coalition (SPARC), who created the OER metadata Rosetta Stone to help translate between different metadata schemes and to suggest which metadata should be required in all repositories (SPARC 2021). However, it is unclear the extent to which user research informed these efforts, and future work to improve OER discovery and curation would benefit from concrete information about user needs and priorities, presented in a way that is accessible to and useful for designers.

Personas

Originally developed by Alan Cooper, personas are descriptive models that aim to precisely convey the goals and needs of a group of users (Cooper et al. 2014, 61). While personas are fictional depictions of individual users, they are grounded in research and serve as archetypal, aggregate composites based on groups of users with similar behavior patterns that preserve individuals’ anonymity. According to Cooper and colleagues, personas aid design by having a narrative structure and resemblance to a real person who is more easy for designers to relate to. Their use is based on the premise that design is more effective when it aims to meet the needs of specific individuals whose goals are representative of a larger group of users, rather than when it tries to accommodate long lists of disconnected needs from disparate users (Cooper et al., 2014, 62–64). Aldin and Pruitt describe the benefits of personas as “making assumptions and knowledge about users specific, creating a common language with which to talk about users meaningfully” as well as allowing designers to focus on smaller subsets of specific users and creating empathy for those users (Aldin and Pruitt 2010, 1). A Delphi study by Miaskiewicz and Kozar (2011, 423) found that, with moderate consensus on ranking, experts identified the top ten benefits of persona use as:

1. focusing product development on users and their goals;
2. prioritizing product requirements and determining if the right problems are being solved;
3. prioritizing the audience;
4. challenging organizational assumptions about users;
5. avoiding design that references the designer rather than the user;
6. providing a clear picture of customer needs and context to guide design decisions;
7. aiding in achieving design agreements and clarifying user goals to interested parties;
8. engaging, unifying, and educating those who are more distant from users and user research;
9. creating empathy for the user; and
10. stimulating innovative thinking for solutions to user goals and challenges.

Personas are increasingly being used in library contexts to guide user-centered design of services (Brigham 2013; Koltay and Tancheva 2010; Tempelman-Kluit and Pearce 2014; Zaugg, et al., 2016; Sundt and Davis 2017). Sundt and Davis relate how at Utah State Universities, prior to use of personas, website design was largely driven by “internal preferences and needs” with any consideration of users limited to “anecdotes or assumptions … with little agreement on how to evaluate or prioritize their assumed characteristics” (2017). Their use of research-grounded personas supported taking a more user-centered approach to design, and gave those collaborating on the design process a common frame or reference to discuss user needs. Lewis and Contrino (2016) describe how personas of distance-education library users were used to better bridge gaps between designers’ and users’ mental models of digital libraries, resulting in improved, data-driven design decisions about digital-learning objects and website design. Some personas are even used by libraries to improve targeted marketing to potential patrons (Thompson, Eva, and Shea 2017). Libraries have also found persona research to be useful outside of the specific institution in which it had been conducted. Zaugg and Ziegenfuss (2018) found that personas of library patrons developed for one academic library were applicable to a second library, demonstrating the potential universality of some library personas in similar contexts.

While personas have the ability to aid the design process, one significant drawback is that, as archetypes, they are particularly vulnerable to stereotyping (Turner and Turner 2011; Hill et al. 2017). Sundt and Davis (2017) recommend ensuring persona descriptions are solely based on rigorous research using real user data in order to mitigate the risk of researcher or designer bias informing persona development. Brummer (2021) recommends building personas by focusing on behaviors rather than demographic characteristics that lend themselves to stereotyping. While such strategies can be helpful, it should be noted that even when demographics are left out of narrative descriptions, designers can take cues from photographs and names and be influenced, consciously or unconsciously, by assumptions about gender, age, ethnicity, or other perceived demographic characteristics. There is also the risk, when personas with demographic cues are attached to professions, of reinforcing representational inequities.

Balancing the importance of names and faces in personas to making users concrete and eliciting empathy in designers, on the one hand, with risks of stereotyping based on visual or name cues, on the other, is a complex issue. Visser and Stappers (2007) note that research and theory in the fields of visual communication and media studies suggest that images of users can stimulate empathy, support creativity, build trust, support storytelling, and support recall through anchoring. Their research on the use of images in personas confirmed that photos in combination with names received the most attention, and that fictively attributed photos worked better than sketches. Turner and Turner (2011) reference using multiple photos for each persona as one potential method of mitigating the problem of stereotyping through images; however, Salminen et al. (2018) found that using photos of multiple people for personas can create confusion and lower persona informativeness. Other methods of stereotype mitigation include having multiple and diverse participants in persona development, educating designers about the social science of stereotyping, and incorporating activities into the design process that surface developers’ implicit biases and assumptions in order to consciously mitigate them (Turner and Turner 2011). Some researchers may choose to leave photographs and names out of persona descriptions altogether due to stereotyping concerns. For this project, persona narratives were based on interview findings and did not include demographic details. We elected to use photographs and names for each persona to make persona memorable and inspire more empathy in designers and to support design storytelling However, in selection of openly licensed[1] photographs for the personas, we also made a deliberate attempt to not reinforce specific stereotypes as to the gender, age, or ethnicity of university faculty and librarians, though we acknowledge the role that unconscious bias can play in the selection process.

Methods

Between December 2020 and January 2021, in-depth interviews were conducted in two phases with 35 librarians, staff, and faculty. The first phase consisted of interviews with 30 librarians, digital- or distance-education staff, and faculty from institutions affiliated with LOUIS, OhioLink, VIVA, PALNI, PALCI, and DigiTex. The second phase featured interviews with five collection specialist librarians, four of whom worked at the consortium level.

In phase one, an initial pool of potential participants were identified from a list of recommendations supplied by representatives of LOUIS, OhioLink, VIVA, PALNI, PALCI, and DigiTex. Consortia partners were asked to provide a list of faculty and library staff from both large and small two-year and four-year institutions, that represented a range of subject-area expertise, and that represented diversity in race/ethnicity and gender. A survey was then administered to recommended participants that included questions on discovery and metadata practices and needs. Sixty faculty, librarians, and staff responded, and a subset of 23 participants reporting a moderate to high level of experience curating OER were selected to participate in interviews. To increase the diversity of the institutions represented in the interview pool, an additional nine participants who did not take the survey were invited from Historically Black Colleges and Universities (HBCUs), Hispanic Serving Institutions (HSI), or schools where more than 50% of the students were Pell grant recipients. The final pool of interviewees contained 18 faculty members representing a diversity of disciplines (Table 1), 10 librarians, and two staff members who worked with OER in a distance or digital education role. Fourteen of the 30 participants represented schools where more than 50% of the students received Pell grants, five worked at an HBCU, and four at an HSI. Twenty-one participants (70%) came from public institutions (Table 2).[2]

Each of the thirty faculty, librarians, and staff members participated in ninety-minute Zoom interviews consisting of four components:

• open-ended questions about their perceptions, attitudes, processes, and experiences around OER curation;
• a think-aloud activity where participants were asked to screen-share and demonstrate their OER search processes;
• a card sorting activity in which participants sorted nineteen metadata items corresponding to metadata concepts already available in the OER Commons platform into the categories of primary consideration, secondary consideration, rarely useful, or it depends; and
• a stimulus response to mockups of new approaches to presenting metadata.

A qualitative, inductive, thematic analysis of interview data was conducted in order to identify shared discovery and curation goals amongst participants, and shared needs and pain points associated with these goals. Similar goals and associated needs and pain points were clustered to create two faculty user personas and one reference librarian persona. Results of the card sorting activity and the think-aloud were used to determine which metadata were used to support the discovery and curation goals associated with the different persona, when they were used, and how useful they were. Written user persona descriptions were then developed that included a fictional name; a photograph; a role; a quote summarizing the essential motivation of the user (created from composite interview data); a narrative that gives insight to each user’s goals, needs, and pain points; and a set of metadata items important to that user. As noted above, persona photographs were sourced from the public domain, and beyond the attempt to have demographic diversity represented in the five persona pictures, there was no deliberate assignment of a particular photo to a particular persona.

In addition to the personas, interviews were used to construct a sequential model representing users’ OER discovery– and curation–decision-making processes. This model includes essential questions asked and metadata used at each stage of a user’s OER search.

As the initial sample of ten librarians were found to contain too few who dealt with curating and maintaining library OER collections, as a second phase, five collection librarians were recruited through grant partners and ISKME’s existing OER librarian network. These collections librarians represented four of the six consortiums collaborating in the research grant, and one consortium in the northeastern US that was not part of the grant but that does have an OER commons microsite. Collections librarians participated in ninety-minute semi-structured interviews which consisted of open-ended questions relating to their role and responsibilities, OER curation approaches, challenges, wants and needs, and metadata preferences. Two additional collections librarian user personas, one whose goals focused on building and maintaining OER collections and one whose goals focused on aligning OER to specific courses, were developed based on these interview data.

Findings

A total of five curation personas were developed from the interview data, detailing OER discovery goals, needs, and pain points, and metadata utility: Faculty Textbook Replacer, Faculty A La Carte Curator, OER Reference Librarian, Collections Maintenance Librarian, and Course Redesign Support Librarian. Below, we present a description of these five personas, each of which represents an amalgamation of similar users. We also include the five persona slides, shared in the form they were presented to the ISKME design team (Figures 1 through 5).

Faculty Textbook Replacer

The Faculty Textbook Replacer persona (Figure 1) embodies a faculty user who simply wishes to replace a commercial textbook with an OER textbook. They typically teach introductory, lower-level classes in a subject area where a textbook is traditionally used, such as introductory STEM courses or an introductory psychology course. They are often motivated in their OER search by mandates or incentives for OER conversion and the desire to make textbooks more affordable for their students.

If the Faculty Textbook Replacer was accustomed to ancillary materials such as presentation slides, problem sets, and test banks being packaged with their commercial textbook, then they similarly want to be able to find these resources packaged together in OER. Faculty Textbook Replacers often must be ready to prepare to teach a class with little notice and are often searching without the support of a librarian. Consequently, the Faculty Textbook Replacer needs a quick and easy way to search for OER. A Faculty Textbook Replacer typically searches by keyword or by subject and prefers to look in repositories that are focused on textbooks, like Lumen Learning or the Open Textbook Library. They struggle when there are broken or outdated links, or when a search returns duplicates of the same material and insufficient resources. They would benefit from metadata that shows them if a resource is aligned with their class content, like a table of contents, and metadata that tells them if a resource comes with ancillary materials. They would also like to have information that allows for rapid evaluation of the trustworthiness or pedigree of a resource, such as faculty reviews and provider metadata, as well as information that shows how easy the resource is for their students to use, like material type, format, and accessibility metadata.

If you’re going to take away my commercial text, I need a replacement with the same features

I get that education access is important, and I support my community college’s mandate that we shift toward using OER. But finding quality open materials isn’t always easy.

What’s ‘quality’? For me, it’s the resemblance to a commercial textbook. The presentation has to be professional. There has to be a natural progression of the content—an internal consistency. It has to have test banks and ancillaries like a commercial textbook. And it has to come packaged as one thing.

I don’t have time to cobble together bits and pieces and adjust them so that they integrate. That’s not workable—especially when we’re parachuting-in an adjunct at the last minute. I need a single resource I can use to replace a commercial text, and sometimes it’s not easy to find.

Once I select my OER, I want to import everything into a course manual/ companion so that I can post it into my LMS to prevent students from getting derailed by external links and clicks.”

Faculty A La Carte Curator

In contrast to the textbook replacers, the Faculty A La Carte Curator (Figure 2) takes a more exploratory approach, piecing together different OER resources of interest. They are more likely to be teaching upper-level classes in subject areas that don’t lend themselves to textbooks, such as literature or film, so will search more diverse repositories like OER Commons, or even non-OER sites they know have strong resources like the Smithsonian. Because of this, they would benefit from a way to “save” items of interest to go back to later or easily compare saved items found across repositories. They are more deeply invested in OER and may want to remix a resource, and so knowing the license type and the limitations of use are very important. They have a higher tolerance for research frustration than the textbook replacer, but do complain about needing to search multiple repositories, metadata that is inconsistent across repositories or insufficient to their needs for assessing resources, lack of available resources, and the complexity of managing the many potential items in a search process. Similar to the Faculty Textbook Replacer, they are aided by a table of contents to determine alignment to their course needs, user evaluations to determine quality, and accessibility metadata to ensure resources can be used by students with learning accommodation needs.

I take pride in customizing my courses each year with new and topical resources that bring out the best in me as a teacher, and in my students as learners.

I’ve never liked commercial textbooks much. Teaching from the same dense book year after year is not a recipe for student engagement—or my own. I’m always looking for new OER; not just when I’m planning my courses, but all year long.

It’s fun for me to go down the rabbit hole—finding things I haven’t seen before and getting ideas. Librarians have helped me become a better searcher, but probably there’s more for me to learn. I want the OER movement to transform teaching—not just by making more stuff available, but by creating a kind of interactivity that didn’t exist before.

Once I select my OER, I want to save and organize items so that I can integrate them later. I then want to sequence items from a breadth of sources and resource types so that I can create a custom course in my LMS for my specific needs.”

Our faculty interview participants were divided nearly evenly into Faculty Textbook Replacers and A La Carte Curators, though it should be noted that our sample of participants may have a self-selection bias towards A La Carte Curators because of their pre-existing tendency toward deep engagement with OER.

OER Reference Librarian

The OER Reference Librarian (Figure 3) typically needs to be able to support faculty in their OER curation process by finding resources on request. They also need to be able to teach faculty the skills to conduct searches themselves. OER reference librarians are skilled at more sophisticated discovery techniques such as using known aggregators of a desired material type, using limiters, and entering the ISBN of a commercial textbook into a search engine that will suggest suitable open resources that are related. They will first use their library’s licensed collection, and then may search in OER mailing lists for threads about the subject area in question. As their job is to search multiple repositories, version or date-last-updated metadata is needed to help the reference librarian keep track of duplicates and know when changes have been made to a resource. Moreover, as they may not necessarily have the same level of subject-matter expertise as faculty, the subject and table of contents are critical to supporting them in finding material that match course needs.

A librarian also is a copyright expert, so license type is needed to confirm that a resource is OER and to understand the limitations on use as there are various types of open licenses that afford different permissions. Like faculty, librarians also need accessibility metadata to ensure inclusive access. Also like faculty members, reference librarians want quality indicators. This includes user evaluations, but for the reference librarian it also includes the provider and what institution may have vetted the resource. Reference librarians are typically aware of which OER providers and curated repositories have a reputation for quality assurance. A pain point for reference librarians are search terms that don’t turn up sufficient relevant content: the reference librarian needs a simple search system because it needs to not just work for them, but also for the faculty they train.

I enjoy searching for OER to meet individual faculty needs, but it could be less complicated.

I’m managing and troubleshooting electronic resources like databases and eBooks on our myriad platforms. A good part of my work relates to OER, and faculty reach out to me for support with searches, which sometimes means guiding them through a search and other times means doing the search for them.

I am an evangelist for OER, and a competent curator, but even for me the process can be complicated. As the OER movement evolves, I’d like to see a process that is more efficient and simple—both for me and for the faculty—whose buy-in we need for the movement to really grow.

Mira’s metadata needs:

1. Material Type and Format
2. Vetted by and Provider
3. Data Updated/Version
4. User Evaluations
5. License Type
6. Accessibility
7. Subject and Table of Contents

Collections Maintenance Librarian

The Collections Maintenance Librarian (Figure 4) is focused on building out their libraries’ curated OER collections. They have trusted repositories they go to that they know to have rigorous curation, where they can cherry-pick content that aligns with their faculty needs. Therefore, like the resource librarian, metadata on providers and who the OER was vetted by are important to their quality evaluation, and license type, accessibility, and table of contents are important metadata points for both finding what faculty need and making resources added to the collection findable for faculty. The Collections Maintenance librarian would also benefit from media format metadata that indicates, for example, if a resource is available as downloadable file, not just as web-based/HTML material.

A Collections Maintenance Librarian needs to be able to build out collections to address gaps and create breadth in their collections, and to know when a resource has been updated so that they can keep their existing collections up to date. Therefore, date and version are important metadata, and lack of ability to search across systems, inconsistencies in metadata such as differences in institutional field naming, and duplication of resources in a single repository or site are pain points. For the Collections Maintenance Librarian, being able to both share their own curated resources and access those of other libraries would be a huge boon.

I focus on curating for breadth and on supporting faculty discovery of OER in my OER collections.

I work to build out our existing collections of OER so that I cover the greatest breadth of subject matter possible, and organize materials so it’s easy for faculty to identify what they need. I typically curate from collections that I know and that have indicators of quality, like faculty reviews.

I often find that there is a lack of adequate controlled language for subjects in the higher education space, and that there’s an overall inconsistency in metadata across repositories, which slows me down. Because I think about discoverability, I’m concerned about the lack of metadata to handle the varied types of resources that faculty search for, and that OER aren’t embedded into the discovery systems they use.

Big picture, I’d like to be able to efficiently leverage the curation work of others (e.g., through collections-level metadata), and to also to share the curation work I’ve done to benefit the wider OER community.

Course Redesign Support Librarian

The Course Redesign Support Librarian (Figure 5) typically has a role that involves mapping OER to specific courses for the purposes of textbook replacement or for larger state initiatives. This work aids faculty in transforming courses with commercial course materials into OER courses.

Similar to other librarians, the provider and who the resource was vetted by are important metadata, as high-quality textbooks and ancillaries are often found in trusted repositories. Material type and format are also important metadata to have early in the search process.

Like the Collections Maintenance Librarian, the Course Redesign Support Librarian feels frustration at not being able to leverage work being done by other library consortia, effectively meaning that Collections Maintenance Librarians across the country are duplicating work and starting from scratch with mapping efforts. The Collections Maintenance librarian would like to be able to search and share resources and metadata across trusted curated repositories while also being able to easily distinguish duplicate records from updated or remixed versions of the same original resource. Faculty reviews would also be helpful for course mapping and quality assurance. They would additionally benefit from accessibility, subject, and table of contents metadata to ensure alignment with course and accessibility and accommodation requirements.

If other states and consortia are also aligning OER to their courses, why can’t I leverage that?

I support the curation of OER for textbook replacement and course redesign—either as part of individual faculty projects or as part of broader initiatives for mapping OER to state-level course requirements.

Sometimes I cherry pick materials in gap areas, and other times I curate with a lens toward mapping OER I find to as many courses as possible within a discipline. I really need a way to increase my success in finding hard-to curate-for upper level courses, including enhanced metadata to help in aligning materials outside of my area of expertise. I also want more detailed metadata that can help faculty discover the materials they need (e.g., accessibility metadata, more nuanced material type metadata, etc.).

I really wish I could more easily leverage and contribute to the curation work of other consortia, for example through a master record where participating libraries can access shared metadata, and add to it, as well as download and integrate it into their local records.

Similarities across personas

While the above descriptions indicate distinct differences in these five types of users, a cross-persona analysis revealed that OER faculty and librarians also share common motivations, goals, and pain points. Student success, decreasing student costs, and increasing student engagement are fundamental motivators for the users’ engagement with OER. Faculty and librarian users also share the goal of finding high-quality, up-to-date, accessible, comprehensive, and well-aligned content. Shared pain points include the difficulty of searching multiple repositories, getting duplicate results, and finding ancillaries that match with texts (Table 3).

Of particular note is the presence of OER accessibility as critical metadata for four out of the five personas. For many users interviewed, knowing that course materials were accessible (i.e., knowing that materials met basic WCAG requirements as is required under the ADA, or would be usable by students in their courses with specific accommodation needs) was important not just because materials that are already accessible reduce high time and monetary costs associated with accommodations, but also because it constitutes a fundamental equity issue. One participant, who categorized accessibility as “primary” metadata explained that you can’t have equity without access. Another participant said, on the importance of having metadata that identifies accessible OER, “I think it [material accessibility] is good and it’s the right thing. But that doesn’t mean it’s always the easy thing.” One participant, who said that they had worked in disability services, suggested that the potential for having resources that are already accessible is an argument in favor of using OER in the classroom rather than traditional textbooks. Research participants typically discussed accessibility metadata needs in terms of information about whether or not the resource met standards for learner differences access. “I’m looking for Universal Design,” shared one participant, referencing CAST’s Universal Design for Learning Guidelines (CAST 2018). Other participants discussed accessibility in conjunction with the importance of material format metadata. These users wanted to know if the format was easy to download for students who do not have consistent internet access and whether it was in a format that allows for revision by instructors and/or students.

A model of curation-decision points

In addition to the curation persona development, in order to better understand user metadata needs, interview data was used to develop a model for sequential curation-decision points and metadata needs at those points. This model reflects the fact that users commonly seek to answer four essential questions in assessing resources in their OER search process (Figure 6):

1. What is this resource?
2. Is the content in this resource a fit for my needs?
3. Is this resource a quality resource?
4. How easy will it be to use this resource?

At each stage in their resource-assessment process, users found different metadata to be valuable in helping them eliminate unsuitable OER in their search process. The first stage involved eliminating resources that were not the correct type of resource. Most useful in this assessment are metadata on the material type. As one participant explained, “That’s primary. Is it a reading? Is it a textbook? Is it a video [lesson]? That is very helpful to me, because then I don’t have to dig in deeper and see what it is.” For those users for whom the lack of ancillaries is a dealbreaker (Faculty Textbook Replacers, for example), knowing up front whether a textbook has them also allows for quick elimination of unsuitable resources at this early stage. Resource format (i.e. whether a textbook has interactive assessments like those created with H5P or plain-text summary questions like you would find in a traditional textbook) was considered lower utility metadata at this early stage, but it did help those users who had strong preferences for or aversions to certain formats.

The second stage in the curation process was determining the fit of resources for specific needs. Title, subject area, learning level, and description were considered essential fundamental information to determine fit. The table of contents was also considered an efficient way to get a good sense of whether the resource was well aligned with course needs. Tags, however, were considered less useful metadata in assessing fit.

After assessing fit, users looked at the quality of resources. Here, faculty evaluations were considered to be of much higher utility than number of downloads or adoptions. As noted above, users often gauged quality by who vetted the material and what provider distributed the resources.

Finally, at the end of the search process, users considered ease of use. Faculty who did not eliminate certain formats at their first decision point took this opportunity to consider resource format. Also important at this stage were metadata on accessibility and license type, which determines whether a resource can be remixed and revised.

Metadata to answer “What is this resource?” include: material type; ancillaries included?; resource format.

Metadata to answer “Is the content a fit for my needs?” include: title; description; subject area; learning level; table of contents; date updated / version; alignment; tags.

Metadata to answer “Is it a quality resource?” include: user evaluations; provider; vetted by (and vetting institution); author; downloads / adoptions.

Metadata to answer “How easy will it be for me to use this resource?” include: license type; accessibility; resource format.

While knowing the order in which different metadata are likely to be used in a discovery process is potentially helpful in optimizing the OER search experience for users, it is important to remember that the persona findings demonstrate that relative utility of metadata is variable depending on who is looking and what they are looking for.

OER Commons Design Application of Persona Research

To better illustrate how personas can be applied to design, we offer the example of how ISKME has planned a multi-phased approach to designing solutions for higher-education users of OER Commons and partner microsites in response to the needs and pain points identified in the faculty and librarian personas. The initial phases of this endeavor involve responding to the following findings represented in the cross-persona analysis:

• Goal 1: find comprehensive content that covers all the topics in a course
• Goal 2: find content that has been positively reviewed and used by people with trusted credentials
• Goal 3: find the most up-to-date content available
• Pain Point 1: too many places to search and many of them contain the same materials
• Pain Point 2: difficult to tell the difference between versions (if duplicate is the same textbook or an update or remix)

To address these user needs, ISKME is in the process of developing an exchange that will help partner institutions share and search for OER materials across digital library platforms. The first milestone in the exchange development process is providing a strong, centralized, cross-platform search-and-discovery user experience, to address Pain Points 1 and 2 above. Librarians will be able to select specific collections from their local microsites and hubs to share with the entire community, as well as search for content shared by others. This will give librarians access to resources that have been authored and curated by trusted institutions, reduces duplication of effort in curating and applying metadata to the same OER content over and over again, and increases visibility and usage for novel content created and shared by faculty in local repositories. Furthermore, in addressing Pain Point 2 above, it was important to ensure search results are deduplicated and that metadata from multiple repositories are merged to provide clean search results with enriched metadata. Therefore, ISKME is developing automated ways to recognize when the same resource is cataloged on multiple sites by matching elements such as source URLs, titles, abstracts and descriptions, or the text itself. This not only allows users to more easily navigate search results when the same source is indexed across multiple microsites; it also facilitates the amalgamation of metadata and user reviews, and allows users to know how many institutions have chosen to include a resource in their catalog, ultimately giving users valuable information for course-alignment and quality-evaluation assessments. The OER exchange also will allow librarians from different consortia to leverage each other’s metadata, tagging, and alignment information, saving time and resources.

ISKME’s second milestone design goals will build on this work, making it easier for faculty and librarians to accomplish Goals 1, 2, and 3 by focusing on features that provide:

• an easy process to allow librarians to index collections from exchange partners in their local repositories, including automated metadata mapping between different local taxonomies, so that faculty and librarians have access to the full breadth of content created by the community, tagged and indexed in a way that is locally relevant
• the ability to subscribe to automated collection updates from exchange partners so that faculty and librarians consistently have access to new content, new versions, and new metadata as they are published
• metadata enrichment across the OER exchange community to improve substance, utility, and clarity, such as:
  • improving the design of content-review features to enable and encourage users to leave helpful reviews, including star ratings and evaluations on dimensions such as comprehensiveness and quality of pedagogical design
  • improving course-alignment metadata by having examples of syllabi of courses that have used a textbook or an OER resource, and examples of key topics included in the resource

Additional future milestones will focus on developing more sophisticated features that are capable of addressing the following pain points:

• Pain Point 3: difficult to match ancillaries to core texts
• Pain Point 4: difficult to tell how accessible content is

These improvements, guided by curation personas, will allow higher-education faculty and librarian users to more easily find and evaluate the fit of resources and related ancillaries in the early stages of their OER search, as well as supporting Collections Maintenance Librarian and Course Redesign Librarian user types in the specific work they do in cataloging and collection building.

Discussion

User-experience research is by nature rooted in the ability of a specific service provider to respond to their users’ needs through design solutions for their particular service context. However, our research sought not just to understand how users interacted with ISKME’s OER digital library, but to more broadly understand higher-education faculty and librarians’ OER discovery and curation process, including goals, needs, and pain points. As such, the personas and curation decision model developed in this study have the potential for broader applications in the OER field.

The identification of three types of librarian users with both distinct and overlapping discovery and curation goals is particularly important given the critical role librarians play in OER adoption and sustainability. The librarian personas can help designers who are working on discovery solutions understand and contextualize different librarian needs. ISKME designers applied this by noting first where librarians’ goals, needs, and pain points overlapped with faculty goals, needs, and pain points in order to prioritize design choices that would have an impact on a wide variety of users, while also using specific librarian personas to inspire creative design choices to meet the needs of librarians doing collections and course-alignment work that affects a large number of OER users. Other OER libraries and repositories could similarly use OER curation personas to help set design priorities and to increase awareness of the specialized needs of librarian users whose work has a high impact on OER discovery at institutional, consortial, and state levels.

Faculty personas are also useful in defining two distinct types of users with very different OER goals. The Babson Research Group’s 2017 OER survey indicated that the most frequently cited reasons for faculty not adopting OER in the US were “difficult to find what I need” and “lack of resources for my subject” (Seaman and Seaman 2017, 30). However, personas revealed that “what I need” and “resources for my subject” are very different for Faculty Textbook Replacers versus A La Carte Curators, meaning that those seeking to alleviate this barrier must consider what search processes, metadata inclusion, and search-result displays work for two very different curation approaches based on different needs. At the same time, the two faculty OER curation personas tell us what these users have in common, such as a shared desire to find resources in one place and to have information about resources’ tables of contents, material type, accessibility information, and quality in the form of user reviews. Given that only two types of faculty OER curators emerged from our research, optimizing searches for these two users could meet the needs of the majority of faculty users, not just in OER Commons and partner microsites, but in other OER repositories catering to higher education faculty.

These five personas and the curation-decision model also have the potential to help the larger OER community in informing broader initiatives not attached to particular library or repository design projects, such as the ongoing efforts to improve and standardize OER metadata. For example, looking at what metadata is important to a variety of users, or essential for a particular user, can help guide decisions about what metadata should be universally requested or required when an OER is submitted to a digital library platform or discovery service. Information about how and when metadata is used by different users can also guide choices about what automated metadata searches we should develop and how we can best rank search results. The curation-decision model also generally suggests areas that can be strengthened in the visual display of OER metadata by highlighting what metadata is used at what point in the OER discovery process.

This research also helps illuminate where best practices for OER discovery might already be in place and where there is the need for future efforts in OER discovery research to support best practices. For example, the importance of ancillaries to various user personas and the pain point of not being able to tell the difference between versions of the same OER confirms the importance of the OER Metadata Rosetta Stone’s inclusion of “is ancillary” and “has ancillaries” as well as “edition statement” as core metadata elements. However, these fields are currently relegated to an “optional” classification, because processes for collecting ancillaries and versioning data are nascent in the field as a whole. Our user research demonstrates the importance of allocating more research and implementation resources toward making the collection of this information part of standard OER development processes.

The primacy of accessibility metadata as it pertains to student learning accommodations to nearly all users suggests that filtering by the presence of accessibility information, as, for example, can be presently done in both the OER Commons and the MERLOT platforms, should be standard across all repositories, alongside information about alignment to accessibility standards such as the Universal Design for Learning Framework. However, ISKME’s own internal UX research (in partnership with CAST, the nonprofit organization that developed the Universal Design for Learning Framework) has demonstrated that current accessibility metadata schemas are not understandable or usable by the average instructor or librarian. Future work in this area should focus on the implementation and testing of tools and processes that more effectively tag and describe content to enable OER creators and users to understand how technical features of a digital learning object impact accessibility for students with different accommodation needs.

Finally, the importance of quality assurance in faculty and librarians’ curation processes suggests that OpenStax’s sharing of star reviews, user reviews, number of adoptions, and clear articulation of peer review status, or MERLOT’s use of a “Quality” box that collects all available quality indicators for each resource, might be models for the type of quality information that could be helpful to users. However, given our findings showing that download and adoption data is less useful to users than faculty evaluations, further user research is likely needed to ascertain if available indicators are meaningful to users and sufficient for their quality assessment needs, even when they are more prominently displayed.

The strength of UX personas is that they put users first in the design process. For those OER sites that do not have the resources to engage in their own user research, personas provide a design starting point that is grounded in user experience rather than designer assumptions. They also provide more robustness of understanding and clarity about the different types of faculty and librarians who do or potentially could leverage OER. Librarians and faculty are not monolithic user groups, and understanding nuanced differences between the goals, motivations, and pain points of different types of users is important in ensuring the success of OER initiatives more broadly. For example, as noted above, the importance of encouraging faculty to submit and openly license ancillaries along with their OER Textbooks is clearly illuminated in our research, but this had not been previously prioritized as part of programmatic efforts that support OER authoring.

However, one aspect of the use of personas that remains unclear is whether or not the use of names and photographs in personas is necessary in order to achieve the design output, and therefore worth any risk of a potential stereotyping effect. While ISKME OER Commons designers anecdotally confirm that personas are useful in creating user stories to support the design process, it was beyond the scope of this research study to confirm whether the photographs and names used in the faculty and library personas supported empathy and memorability, or if they triggered unconscious biases. Demographic characteristics such as gender and race/ethnicity did not overtly play a role in ISKME’s metadata design decisions, and the collaborative nature of the design process—which included feedback from external partners, as well as ongoing, internal organizational conversations around diversity, equity, inclusion, accessibility, and anti-bias—ideally helped minimize the risk of stereotypes unduly influencing design for this particular project. However, these personas are publically available as OER, and there is arguably a risk that, if they are adopted for use in other environments, that stereotyping could be even more likely to occur. For example, photographs and names may read as reinforcing race/ethnicity and/or gender stereotypes about the demographics of different types of librarian jobs. For this reason, those using these personas may want to consider conducting their own stereotype risk assessment and removing or replacing photographs and names as appropriate for their own design environment.

Overall, we found personas to be a valuable way of organizing and summarizing research-based findings on user discovery needs and challenges in a way that generated a suite of actionable design decisions. This research focused on higher-education librarians and faculty in the US, but UX research to develop personas for students and for users in the K–12 space would be an important next step in removing the discoverability barrier for OER use. Furthermore, the global OER community would be enriched by persona studies in different countries in support of a worldwide approach to improving OER discovery, and ultimately in service to the community’s greater goals of global information access and educational equity.

Acknowledgments

This article is based on a study made possible in part by the Institute of Museum and Library Services (www.imls.gov), under grant number [LG-246327-OLS-20]. The findings and recommendations expressed in this article do not necessarily represent those of the Institute of Museum and Library Services. We would like to acknowledge Lisa Petrides; Nick Lobaito; Sophie Rondeau; Teri Gallaway; Emily Frank; Anne Osterman; Sophie Rondeau; Anna Bendo; Amanda Hurford; Gretchen Gueguen; Judith Sebesta; and Ursula Pike for their contributions to this research.

About the Authors

Michelle Brennan leads product development for Open Access Digital Library Platforms at the Institute for the Study of Knowledge Management in Education (ISKME), where she collaborates across research, UX, engineering, and partner services. Michelle received her Masters in Information Science from the University of Michigan School of Information where she worked as an Instructional Technology Support Librarian. Her research interests include OER discovery and metadata, and the role of open knowledge tools in education in improving opportunities for participation in collaborative spaces.

Selena Burns is a senior researcher at the Institute for the Study of Knowledge Management in Education (ISKME), with expertise in qualitative and mixed-methodology education research. Her interdisciplinary background includes a BA from Wellesley college in Anthropology with a minor in Psychology, an MA in Communication from Stanford, a PhD in Educational Theatre from NYU, and six years working on federally funded STEM education research projects for Heller Research Associates. At ISKME her research has focused on Open Educational Resources in K–12 and higher education.

Cynthia Jimes served as director of research and learning for the Institute for the Study of Knowledge Management in Education (ISKME) from 2006 to 2022. During that time she led large-scale studies of the use of OER and other adaptable learning materials at the K–12 and post-secondary levels to support enhanced teaching and learning. Jimes has a PhD in Information Science from Sweden’s Uppsala University.

Jeff Hecker is a founding Principal at Athena Brand Wisdom. He has over twenty years experience in market and user research, including prototype research, usability testing, persona development, and research to support information architecture development. Jeff has an MBA in Marketing from the Schulich School of Business at York University and a degree in English Literature and Semiotics from the University of Toronto.

Anastasia Karaglani is a research associate at the Institute for the Study of Knowledge Management in Education (ISKME), focusing on quantitative and qualitative data collection, data management, and analyses. She has worked as a research assistant at USC, CRESST/UCLA, and Stanford University conducting quantitative and qualitative research. Anastasia holds a PhD in Educational Psychology from the University of Southern California.

Amee Evans Godwin is a senior advisor at ISKME, consulting on research projects and leading the national GoOpen initiative. She brings more than three decades of experience in applied research and interactive applications focused on learning and collaboration. Prior to joining ISKME, Amee worked as a consultant at Interval Research and in strategic market research at high-tech firms and nonprofits in San Francisco and Silicon Valley. She has a Masters Degree from New York University Tisch School of the Arts in Interactive Telecommunications, and BA from University of Buffalo’s Center for Media Study.