2026 ASEE Annual Conference & Exposition

This full paper presents an empirical research study focused on the development and validation of a new domain-specific instrument to measure perfectionism in engineering education. Perfectionism is widely observed among engineering students, yet instruments that distinguish healthy striving from concern-driven tendencies within this disciplinary context remain limited. This study reports the creation of the AMPERE (Adaptive and Maladaptive Perfectionism in Engineering) scale through a three-phase process: (I) expert review with two domain experts to establish content validity; (II) cognitive interviews with seven undergraduates to refine item wording and interpretability; and (III) cross-sectional survey administration at a large R1 engineering college.

After rigorous screening of survey responses (N = 317), confirmatory factor analyses compared a theory-driven 20-item baseline against reduced variants. A concise two-factor structure, Model C (17 items), demonstrated the best balance of fit and parsimony (CFI₍robust₎ = 0.822, TLI₍robust₎ = 0.795, RMSEA₍robust₎ = 0.094, SRMR = 0.086, with the lowest AIC/BIC) and a near-zero interfactor correlation (r ≈ .05), indicating empirically distinct adaptive and maladaptive dimensions. Internal consistency was high for both factors (α = .89 and .91). Item loadings ranged from .54–.83, and multigroup comparisons suggested configural stability across gender and academic level.

Findings clarified that adaptive perfectionism—characterized by personal standards, organization, and self-directed striving—was distinct from maladaptive perfectionism, which reflected doubts about actions, external pressures, and fear of evaluation. AMPERE Maladaptive correlated strongly with APS-R Discrepancy (r ≈ .81) and MPS SPP (r ≈ .57), whereas AMPERE Adaptive correlated with APS-R Standards (r ≈ .54) and MPS SOP (r ≈ .41). Criterion-related analyses showed that maladaptive perfectionism predicted higher decision avoidance (r ≈ .22), while adaptive striving related negatively to avoidance (r ≈ –.16). Together, these results provide robust psychometric evidence for AMPERE’s validity and highlight that concern-driven perfectionism, not high standards per se, is linked to decisional rigidity.

This work advances measurement precision in engineering education research by distinguishing healthy striving from maladaptive self-criticism. Practical implications include using AMPERE for early identification and advising, informing Counseling and Psychological Services (CAPS) referral and psychoeducation programs, and supporting mentoring models that balance high standards with self-compassion. Future studies should examine measurement invariance across institutions, explore longitudinal predictive power, and pilot an interactive web- or mobile-based AMPERE feedback tool integrated with campus decision-skills resources.