Instructor Presence Stabilizes Performance Distributions in Digitally Scaffolded CPR Training

Abstract Background: Digital feedback systems are increasingly integrated into school-based cardiopulmonary resuscitation (CPR) training. Prior research has largely emphasized mean performance outcomes, whereas assessment theory suggests that distributional characteristics—particularly lower-tail reliability—may be equally relevant in safety-critical competence contexts. Purpose: This study examined whether different instructional configurations under digitally scaffolded CPR training—peer-led and adult-led instruction—differ in distributional properties of performance, and how these formats compare to a minimal-scaffolding digital-only configuration. Methods: A quasi-experimental comparative design was conducted across three conditions: peer led instruction (n = 342), adult-led instruction (n = 69; individual-level subsample), and digital-only training without instructor guidance during assessment (n = 23). Mechanical CPR performance was assessed using QCPR-enabled manikins generating composite percentage scores. Non-parametric tests were used to compare distributions and frequencies of catastrophic low-performance cases (<50%), operationalized as lower-tail competence failures. Results: Mean performance was comparable across groups (Peer: 84.45%; Adult: 83.54%; Digital only: 82.52%). However, distributional differences emerged. Catastrophic lower-tail failures occurred exclusively in the digital-only condition (3/23; 13.0%) and were absent in instructor present formats. Overall group differences were statistically significant (H = 11.38, p = .003), with pairwise contrasts indicating differences between instructor-present and digital-only conditions, but not between peer-led and adult-led instruction. Conclusions: Although mean performance did not differ, instructor-present formats were associated with reduced lower-tail vulnerability. From an assessment perspective, distributional integrity may therefore complement mean-based interpretations when evaluating instructional effects in safety critical skills training.