Cognitive and emotional predictors of preference for emotionally intelligent speaking assessment among Saudi EFL learners: a structural equation modelling approach

This study examined Saudi English learners’ perceptions of Emotionally Intelligent Speaking Assessment (EISA) through a structural equation modelling approach informed by Expectancy-Value Theory (EVT) and the Control-Value Theory of Achievement Emotions (CVTAE). A total of 754 undergraduate students from Science, Technology, Engineering, and Mathematics (STEM) and non-STEM disciplines completed the validated EISA Questionnaire, developed to explore the emotional, cognitive, and fairness-related dimensions of speaking assessment. Although emotional intelligence has received growing attention in language education, little empirical research has investigated how learners perceive emotionally responsive speaking assessments within the Saudi context, where evaluation remains predominantly product-oriented. The results showed that the perceived importance of emotional expression, fairness, self-confidence, and usefulness significantly predicted learners’ preference for EISA. Mediation analysis revealed that fairness, self-confidence, and usefulness mediated the link between perceived importance and assessment preference, while moderation analysis showed effects of gender, academic field, and speaking proficiency. Rather than treating cognitive appraisals as a distinct construct, the study investigated how learners’ motivational and emotional perceptions jointly shaped their preferences for emotionally intelligent assessment. These findings indicate the importance of integrating emotional and social dimensions into speaking assessment to foster more equitable, authentic, and communicatively relevant evaluation practices.