Factors and Lessons Learned from Moving Beyond Satisfaction Measurement to Learning Outcomes in Enhancing Training Service Quality: Lessons from Mahidol University International College (2019–2024)

Abstract

This research article presents empirical findings and synthesizes new knowledge to address a significant theoretical and practical problem in training evaluation, which is the empirical paradox. This paradox refers to the inconsistency between perceived service quality and learner satisfaction (Kirkpatrick Level 1) and the actual measurable learning outcomes (Kirkpatrick Level 2). The issue has been confirmed by multiple international meta-analyses, which highlight the limited correlation between Level 1 and Level 2, posing a major challenge for training quality management in higher education institutions. Accordingly, the primary objective of this article is to validate the empirical paradox and to synthesize “strategic compensatory factors” in order to develop an outcome-oriented model of training service quality.

This study presents lessons drawn from empirical analysis using quantitative secondary data from nine corporate training programs conducted by Mahidol University International College (MUIC) between 2019 and 2024. It applies an integrated conceptual framework based on the SERVQUAL model to assess service quality dimensions, alongside the measurement of normalized learning gain (Hake’s G) to evaluate learning outcomes at Kirkpatrick Level 2. This combined approach allows for a more comprehensive assessment of both perceived service quality and actual learning effectiveness.

The key findings indicate that although Level 1 scores (learner satisfaction) were consistently high, Hake’s G values showed significant variability (ranging from 0.45 to 0.82), thereby confirming the empirical paradox. A major contribution of this study is the identification of a “strategic compensatory factor,” a theoretical mechanism linking the Assurance dimension of instructors to learning gain through mediating variables such as learners’ self-efficacy and perceived utility. These insights are synthesized into an outcome-oriented training service quality model, accompanied by a balanced Key Performance Indicator (KPI) framework. The model emphasizes the development of AI fluency among instructors as a key driver of training quality and learning outcomes.