SELF-LEADERSHIP AND PSYCHOLOGICAL STRESS AS PREDICTORS OF ATTITUDES TOWARD ARTIFICIAL INTELLIGENCE AMONG PAKISTANI COLLEGE STUDENTS: A MULTI-THEORETICAL EXAMINATION

Abstract

Artificial intelligence is now embedded in routine higher education practice, yet comparatively little is known about the psychological characteristics of the students using it. The present study addressed this gap through four theoretical frameworks: Self-Leadership Theory, Social Cognitive Theory, the Transactional Model of Stress and Coping, and the Technology Acceptance Model. Two hundred twenty-three Pakistani college students completed the Abbreviated Self-Leadership Questionnaire, the General Attitudes toward Artificial Intelligence Scale, and the Stress subscale of the Depression Anxiety Stress Scales (DASS-21). Internal consistency was adequate for all scales (alphas .77 to .86). Self-leadership was positively associated with positive AI attitudes (r = .20, p = .003) but showed no association with negative attitudes. Stress correlated positively with positive attitudes (r = .22, p = .001) and negatively with negative attitudes (r = -.35, p < .001). In the regression models, both predictors remained significant for positive attitudes (R² = .109), whereas stress was the sole significant predictor of negative attitudes (R² = .129). Interpreted through a transactional appraisal framework, this pattern suggests that students under academic pressure may appraise AI as a coping resource rather than as a threat. Self-regulatory capacity and stress appraisal therefore appear to operate through distinct pathways, a distinction with implications for institutions seeking to support constructive student use of AI.