Student-Led Evaluation Identifies Friction in Guadalajara Digital Library

Undergraduate students who use the Universidad de Guadalajara’s Biblioteca Digital encounter a mismatch between the platform’s institutional organization and the way they search for academic material. A five-stage, mixed-methods evaluation found that none of 24 students completed every assigned navigation task successfully. In a separate questionnaire, 93% of 58 respondents could not correctly identify the contents of the online catalog and 84% were unfamiliar with Boolean operators.

The study examined the university’s principal digital gateway to scholarly information, a platform that provides access to more than 117 specialized databases, nearly 90,000 electronic journals, more than 4.2 million electronic books and over 44,000 theses, according to the research article. It also includes institutional publications and a digitized historical collection. Although the archive is extensive, access depends on users being able to interpret its menus, labels, icons and search functions.

The research therefore examined whether the site’s information architecture—the organization, labeling, navigation and search systems that determine how content is grouped and found—creates barriers for the students who account for most of its recorded use. The investigators began with a descriptive analysis of platform access logs. Undergraduate students generated 85.80% of the accesses recorded during the period examined, establishing them as the platform’s primary user profile and supporting their selection for the subsequent studies.

Comparable disaggregated information for academic and research staff was unavailable in the analyzed logs. The project then moved through four additional stages: remote usability testing, a terminology-comprehension questionnaire, closed card sorting and an evidence-based redesign. Each stage was intended to inform the next, combining behavioral observations, semantic assessments and comparisons between institutional categories and students’ own ways of grouping information.

Twenty-four undergraduate students, all between 20 and 22 years old, completed the remote, unmoderated usability test from their own devices. Task success was defined as reaching the specified target screen within the allotted time without requesting external assistance. Although every participant reported prior familiarity with the platform, none completed all assigned tasks successfully.

Click tracking and heat maps concentrated attention on the main navigation bar, internal search engine and informational banners, but repeated interaction with those elements did not consistently lead to successful completion. The results show incomplete task performance among a small sample of experienced users rather than failure by first-time visitors. The behavioral pattern suggested more than uncertainty about where a visible control might be located.

Students trying to navigate through menus frequently experienced delayed completion or task abandonment, after which many turned to the internal search tool. Their expectations often did not match that tool’s behavior, particularly because they approached it with habits formed by conventional web search engines. The authors interpreted the convergence of incomplete task completion, menu detours and mismatched search expectations as evidence of structural and semantic friction rather than simple unfamiliarity with the platform.

The finding does not isolate a single technical cause, but it indicates that the way the library is organized and labeled did not readily support successful navigation for these students. A separate questionnaire completed by 58 students assessed 15 multiple-choice and open-ended questions about labels, technical vocabulary and icons. Comprehension varied sharply across the platform’s access points.

While 65% of respondents could define the concept of a repository and 60% correctly described the content of the Institutional Repository, 93% could not identify the contents of the online catalog. Likewise, 84% were unfamiliar with Boolean operators even though the platform’s search functionality relied on them, and most participants did not recognize icons indicating resource types such as English-language articles, reference material or full text. The results therefore point to both structural and linguistic barriers: students did not always understand what major sections contained or how search features operated.

The platform’s built-in glossary did not appear to provide a route out of the confusion. All 100% of questionnaire respondents indicated that they would search for unfamiliar terms externally through a general internet search rather than consult the glossary. This finding does not establish how long participants spent outside the platform or whether they later returned to complete a task, but it does show that the available help was not the first place respondents expected to find an explanation.

The authors treated this behavior as evidence that technical vocabulary and unfamiliar iconography formed part of the access problem, adding an additional layer to the navigation difficulties identified during usability testing. In a closed card-sorting exercise, 20 different students grouped 30 predefined content elements according to their natural perceptions of categories and relationships. Their choices consistently favored functional and everyday use rather than the library-science taxonomy already used by the platform.

Students placed electronic books, electronic journals, reference works, manuals, theses, dictionaries and encyclopedias together as material. They placed cataloging tools, virtual tools, anti-plagiarism services, reference managers, translators, tutorials, free resources, trial databases and the glossary in a separate tools group. Scientific and editorial production, specialized and multidisciplinary collections, the historical collection, regulations, frequently asked questions and wdg.aula formed a general group, while conferences received their own category.

The exercise indicates that students’ expectations differed systematically from the platform’s existing classification. Using those results, the research team proposed a revised home-page architecture with broader, task-oriented categories. Material resources would be brought together in one menu, academic support and digital utilities would be consolidated under tools, institutional and general resources would form another grouping, and conferences would remain distinct.

The plan simplifies navigation-level names and reorganizes content rather than altering technical labels constrained by institutional cataloging policy. It was presented to the Biblioteca Digital’s coordination team, which discussed the observed behavior, compared the old and proposed structures and responded favorably to the proposal’s institutional feasibility. The response suggests that the redesign was considered implementable, but it is not evidence that the revised structure will improve task success, comprehension or satisfaction.

Several limitations define the scope of the findings. The study was exploratory and descriptive, so its percentages are diagnostic indicators rather than statistically generalizable estimates; no inferential tests were applied to the questionnaire or card-sorting data. Participants were volunteers drawn only from the Centro Universitario de Arte, Arquitectura y Diseño, and no gender data were collected.

Academic and research staff, postgraduate students and other user groups were not represented, limiting generalization beyond the sampled undergraduates. Most importantly, the proposed navigation has not undergone post-implementation validation, so its effects remain untested rather than demonstrated. A dedicated evaluation after deployment would be needed to determine whether the redesign changes outcomes.

The project identifies concrete sources of friction in a heavily used academic digital library while preserving the distinction between diagnosis and remedy. Its convergent evidence—from access logs, observed navigation, terminology tests and card sorting—supports testing navigation categories organized around student tasks instead of relying only on technical classifications. It also offers a replicable model for evaluating digital-library architecture by combining behavioral analytics, usability testing, semantic assessment and card sorting.

Whether the new structure actually makes the Biblioteca Digital easier to use will require deployment followed by a before-and-after usability evaluation using the same task-based protocol, ideally expanded to include postgraduate, teaching and research users.

Subject of Research: User-centered evaluation of information architecture in the Universidad de Guadalajara’s Biblioteca Digital.

Article Title: Impact of information architecture on the accessibility of the UDG Biblioteca Digital: a user-centered usability evaluation

Article References: Anaya-Chora, M., Cisneros-Hernández, A. A., & Ramírez-Mercado, M. D. R. (2026). Impact of information architecture on the accessibility of the UDG Biblioteca Digital: a user-centered usability evaluation. Frontiers in Computer Science, 8, Article 1847791. https://doi.org/10.3389/fcomp.2026.1847791

Image Credits: AI Generated

DOI: 10.3389/fcomp.2026.1847791

Keywords: digital libraries, information architecture, usability testing, academic libraries, undergraduate students, card sorting, UX writing, navigation design, web search, Universidad de Guadalajara, Impact, information