Publication Date

9-1-2026

Document Type

Article

Publication Title

Journal of Academic Librarianship

Volume

52

Issue

5

DOI

10.1016/j.acalib.2026.103320

Abstract

Academic libraries increasingly deploy AI-enabled services such as intelligent search, recommendation systems, and chat-based assistance to support scholarly resource discovery. However, some users remain reluctant to rely on these tools. This phenomenon, often described as algorithm aversion, has been widely examined in healthcare and finance but remains underexplored in library contexts. This study investigates the factors associated with algorithm aversion in academic digital libraries and examines how these factors relate to one another. A two-phase mixed-method design was adopted. First, 41 semi-structured interviews with users from Chinese university libraries were analyzed using grounded theory. This stage produced 58 initial concepts, which were grouped into 25 categories across seven dimensions: algorithmic, platform, individual, task, resource, algorithmic risk and bias, and social influence. Second, the Decision-Making Trial and Evaluation Laboratory (DEMATEL) method was used to analyze the relationships among these factors. The results indicate that search-related factors, including result relevance, accuracy, intelligence, and complexity, are still in the most central positions. System functionality emerged as the most influential factor, suggesting that user resistance may often reflect broader platform conditions rather than algorithm performance. Resource coverage and update speed also showed important causal influence. By contrast, privacy and bias concerns were more likely to appear after users had already experienced dissatisfaction. The findings suggest that algorithm aversion in digital libraries develops through continuing interactions between users and the wider service environment. Efforts to reduce resistance should therefore address retrieval quality, platform usability, and collection support alongside algorithm design.

Keywords

Academic libraries, Algorithm aversion, Artificial intelligence, Digital libraries, Discovery systems, Search systems, User trust

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License

Department

Information

Download

Share

COinS