AI evidence extraction

Scopus AI Among Novice Lecturers: Perceptions, Influencing Factors, Institutional Support

Wei Lun Wong, Lawrence Jun Zhang, Jessie S. Barrot · The Asia-Pacific Education Researcher · 2026

AI-generated evidence extraction, verified across multiple analytical personas. Not a substitute for the peer-reviewed original.

9/10

Relevance

1/4

Quality (LMQS)

Evidence

Citations

0.00

FWCI

This is an AI analysis. Read the peer-reviewed original at the publisher: https://doi.org/10.1007/s40299-026-01099-w

Methodology & findings

Study design

Qualitative explanatory multiple case study design with five semi-structured interviews (30-60 minutes each) conducted via Google Meet in 2025.

Sample

N = 5, 4 groups

Primary method

Qualitative analysis: six-phase reflexive thematic analysis (Braun et al., 2023) using Atlas.ti software. Deductive codes derived from ETAM framework organized into three code families corresponding to three research questions. Inductive codes added for emerging themes not fitting ETAM labels. Two researchers independently coded one full transcript to establish code definitions and boundaries. First author applied coding frame to remaining transcripts with documentation in audit trail. Cross-case analysis examined commonalities and differences; cases not following expected ETAM pattern flagged as potential rival explanations. Reflexive notes kept on researcher assumptions. Screenshots used for triangulation and verification of self-reports.

Main result

The study found that "all participants agreed that Scopus AI reduced the time spent handling large volumes of literature" and that "Scopus AI was perceived as a mechanism for improving coverage." Additionally, "Across cases, collegial influence, concerns about reputation and institutional scaffolding all mattered, but their relative weight differed. Social cues opened the door. Individual judgement, perceived risk and available support determined how far lecturers walked through it." Furthermore, "participants described using Scopus AI to gather and organise sources but consistently drew a line at interpretation, insisting on full-text reading and self-authored synthesis."

Reports effect sizes.

Research paradigm

Constructivist/interpretive qualitative research

Author conclusions

"This study offered a qualitative study of how novice lecturers engage with Scopus AI. They valued it for compressing literature searches and strengthening discussion sections, treating it as a research partner rather than a citation finder. While PEOU was high, trust rested less on understanding the algorithm and more on the curated Scopus database to produce 'confidence with caution' sustained by habitual cross-checking." The authors conclude that "Long-term integration depended on the alignment of integrity-focused policy, workflow-oriented training and incentive-linked infrastructure."

Risk of bias

Selection bias: Small purposive sample of 5 novice lecturers from Malaysian universities; Recall bias: Reliance on retrospective self-reports during interviews; Platform maturity bias: Scopus AI was in beta status during the study; Limited diversity: All participants were permanent lecturers in English language education; Selection bias: purposive sampling of novice lecturers who were already users of Scopus AI; Recall bias: reliance on retrospective self-reports of tool use and perceptions; Platform maturity bias: study conducted during Scopus AI's beta status, limiting generalizability to current versions; Geographic bias: sample limited to Malaysian universities only; Technology adoption bias: all participants had prior awareness and hands-on use of Scopus AI; Selection bias: Purposive sampling of only five novice lecturers who had already used Scopus AI at least once, limiting generalizability to non-users or more experienced lecturers; Recall bias: Reliance on retrospective self-reports of Scopus AI use patterns and perceptions; Platform maturity bias: Study conducted during beta stage of Scopus AI, findings may not generalize to stable/mature platform; Institutional context bias: All participants from Malaysian universities with supportive AI policies; findings may not transfer to restrictive institutional contexts; Social desirability bias: Participants may have over-reported positive perceptions given institutional encouragement of AI use; Sample homogeneity: All five participants were permanent lecturers in English language education with similar academic profiles (1-4 years experience)

Limitations

"Three limitations temper generalisability: a small purposive sample of five novice lecturers, reliance on retrospective self-reports and the platform's beta status." Additionally, the study notes that "Future research should recruit larger and diverse cohorts and compare Scopus AI with open-ended models to isolate the role of database provenance and evolving institutional policies."

Open questions raised

Limited attention to novice lecturers' AI use in their own research writing (versus student-facing pedagogy)
Narrow and descriptive focus of existing Scopus AI studies with limited organizational and career-stage analysis
Lack of ETAM analyses that explicitly incorporate institutional support and provenance-based trust for Scopus AI
Need for larger and more diverse cohorts in future Scopus AI research
Need to compare Scopus AI with open-ended models to isolate the role of database provenance
Limited attention to novice lecturers in AI and writing research—most studies focus on students and general-purpose chatbots rather than faculty research writing

Extracted from: pdfAgreement 62%

Explore related topics

AI-Assisted Discovery Academic Writing Case Study