Abstract

This study investigates the factors influencing perceived quality in tertiary education Recommender Systems (RS) by examining structural relationships among user-centric evaluation constructs. The research developed an evaluation framework integrating objective system data with subjective user feedback to ensure that recommendations aid in selecting educational pathways. Using a Design Science Research approach, the framework underwent three DSR iterations with sample sizes of N = 166, 357, and 719, respectively. Quantitative results from the final iteration revealed several statistically robust significant relationships amongst the evaluated constructs: Perceived Recommendation Set Variety (PRSV) positively influenced Perceived System Usefulness (PSU) (β = 0.36, t = 7.92, p <  0.001) and Perceived Recommendation Quality (PRQ) also significantly impacted PSU (β = 0.36, t = 8.86, p <  0.001). Furthermore, Perceived System Enjoyment (PSE) and Perceived Ease of Use (PEU) contributed to Overall Satisfaction (OS) (β = 0.16, t = 3.03, p <  0.01; β = 0.27, t = 5.28, p <  0.001, respectively), which in turn strongly drove user Interaction (INT) (β = 0.59, t = 19.02, p <  0.001). Additional significant relationships included PSU’s positive effect on Overall Trust in the RS (OTRS) (β = 0.22, t = 5.18, p <  0.001), as well as modest yet significant effects of PRSV and PRQ on OTRS. These findings, derived from exploratory and confirmatory analyses, validate the framework’s effectiveness in capturing behavioural metrics and user perceptions, offering RS developers a robust tool to enhance RS quality in tertiary education.

Download

Share

COinS