Edmund De Leon Evangelista ^{1 *} , Syed M. Salman Bukhari ¹

CONT ED TECHNOLOGY, Volume 18, Issue 4, Article No: ep694

https://doi.org/10.30935/cedtech/19499

Submitted: 02 May 2026, Published Online: 06 October 2026

OPEN ACCESS   73 Views   27 Downloads

Download Full Text (PDF)

Abstract

Generative artificial intelligence (GenAI) is increasingly influencing how students write, create, reflect, solve problems, and engage with learning in higher education. However, the evidence remains divided, particularly regarding motivation, engagement, creativity, perception, self-regulation, and psychologically informed learning design. This study presents a PRISMA-ScR-based scoping review of GenAI in higher education from a motivational and psychological perspective. The final corpus included 195 studies published between 2023 and 2025 within a specified eligibility window of January 2021 to December 2025. Records were dual screened, coded using a structured framework, and analyzed across psychological focus, thematic constructs, GenAI tools, study designs, outcomes, contextual characteristics, self-determination theory (SDT) needs, equity, and evidence strength. The screening agreement was 81.9%, with Cohen’s κ = 0.642. Motivation was the most common primary focus (31.3%), followed by perception (26.7%), engagement (18.5%), creativity (15.4%), and general/other (8.2%). The literature was dominated by ChatGPT and general GenAI systems, while survey-based and observational evidence was far more common than experimental evidence, longitudinal research, or studies reporting objective performance outcomes. Among 155 empirical studies, mixed or conditional outcomes were most frequent (46.5%), compared with predominantly positive findings (21.9%). SDT related evidence was most common for competence (42.6%), followed by autonomy (29.7%) and relatedness (22.6%), although direct measurement was rare. Based on these findings, the study proposes the GenAI engagement framework, a literature informed structure that organizes five recurring components of GenAI supported learning without assuming a fixed sequence or causal pathway. The framework is not presented as a validated causal model, but as a structured basis for future research and instructional design. The findings suggest that the educational value of GenAI depends less on the technology itself than on how it is guided, evaluated, and integrated into meaningful learning activities.

References

Citation