An Enhanced Model for Evaluating Learning Satisfaction in Teaching User Stories: A Confirmatory Factor Analysis Approach

Abstract

Background: Understanding how students experience and perceive learning through user stories provides valuable insights into the effectiveness of instructional design. Huang proposed a learning satisfaction framework in which students’ satisfaction emerges from four factors, namely perceived ease of use (PEOU), perceived usefulness (PU), learning motivation (PM), and overall learning satisfaction (LS). A recent study applied this model to teaching user stories in a software engineering course using Confirmatory Factor Analysis (CFA) and Partial Least Squares Structural Equation Modeling (PLS-SEM) but reported suboptimal model fit, with unsatisfactory SRMR and NFI values, indicating the need for refinement. Objective: This study aims to evaluates an enhanced learning satisfaction model for teaching user stories by identifying key influencing factors, examining their relationships, and assessing construct validity and model fit improvements. Methods: To improve the model, additional theoretical paths were introduced, and survey data were collected from 142 randomly selected software engineering students. The refined model was evaluated using CFA and PLS-SEM, with model fit improvements assessed through SRMR and NFI values. Results: The analysis revealed that PEOU and LM significantly influence learning satisfaction, while PU affects satisfaction indirectly through motivation. These interactions among PU, PEOU, and LM explain how ease of use and usefulness enhance motivation, which in turn increases satisfaction. Furthermore, the enhanced model showed an improved fit compared to the previous version, with SRMR values decreasing from 0.092 to 0.076 and NFI improving from 0.765 to 0.813, confirming better construct validity and overall model fit. Conclusion: The addition of new direct paths from PEOU and PU to LS increased the model’s R² and Q² values, indicating stronger construct validity and better overall fit. The refined structure provides a more accurate representation of how satisfaction is formed and offers a validated instrument for evaluating student learning experiences in teaching user stories within software engineering course. Keywords: learning satisfaction, user story, confirmatory factor analysis, model fit evaluation, PLS-SEM, software engineering education.