Influence of workstation design on intra-rater reliability in sculpting kneading task assessments

Intra-rater reliability is a crucial metric for evaluating the consistency of ratings made by the same individual across repeated trials. This study examines intra-rater reliability for posture assessment during a kneading sculpting task on worktables titled at 0° and 25°, and analyzes how table tilt and other factors affect the evaluation consistency. Thirty-nine experienced sculptors performed the tasks under both conditions in a randomized sequence. One rater assessed upper arm, lower arm and wrist postures across three trials. Postures were assessed using the Rapid Upper Limb Assessment method. Reliability and performance differences using the two table tilts were analyzed using Weighted Cohen’s Kappa coefficient (WCK), Wilcoxon signed-rank test, and Bland-Altman plots. Results showed high intra-rater reliability, with WCK averages of 0.78 (0° tilt) and 0.85 (25° tilt). Bland-Altman plots indicated strong agreement, with mean differences near zero and slightly narrower limits of agreement for the 25° tilt, suggesting greater consistency. Minor systematic bias appeared under the 0° tilt, likely due to higher ergonomic demand. The Wilcoxon test revealed significant differences in posture scores between trials for both tilts (p<0.05). These findings underscore the impact of workstation design on intra-rater reliability, highlighting the value of ergonomic integration in manual tasks. Reliability remained high across tilts, with slight variations linked to ergonomic factors. Future studies should include diverse participants, inter-rater comparisons, and varied task types to strengthen generalizability across manual work scenarios.