Test-retest reliability and practice effects across measures of attention in college-aged adults

Attention is central to neuropsychological assessment, yet the psychometric properties, particularly practice effects and test-retest reliability are less well characterized for common attention measures compared to other domains. Seventy-three undergraduate psychology students (86.5% female; M age = 19.6) completed a two-session battery approximately two weeks apart, including the Paced Auditory Serial Addition Test (PASAT), Auditory Consonant Trigrams (ACT), Digit Span (Forward, Backward, Sequencing, and Total Score), and Conners' Continuous Performance Test-3 (CPT-3). Reliability was excellent for PASAT (r = .86), ACT (r = .78), Digit Span Total (r = .82), and CPT-Commissions (r = .77). Reliability was good for Digit Span Forward (r = .72), Digit Span Backward (r = .71), CPT d' (r = .67), CPT-Omissions (r = .60), CPT HRT (r = .72), and CPT ISI (r = .68). Fair reliability was observed for Digit Span Sequencing (r = .53), CPT-Perseverations (r = .46), and CPT HRT SD (r = .59). Practice effects were largest for PASAT (d = .95), medium for Digit Span Total (d = .55), and small-to-medium for most remaining indices (d = .15-.47), with minimal change for Digit Span Sequencing (d = .08). Findings support cautious interpretation of short-interval improvement and suggest that ACT and select CPT-3 indices may be more suitable for brief retest intervals.