RESUMO
BACKGROUND: The Perceive, Recall, Plan & Perform (PRPP) system of task analysis might be feasible to evaluate occupational performance and information processing strategies for persons with Parkinson's disease (PD). AIM: To evaluate: (1) the random error between raters (inter-rater study), (2) the random error within raters (intra-rater study), and (3) the internal consistency of the PRPP. MATERIALS AND METHODS: (1) video-recorded performance of meaningful activities of 13 Dutch persons with PD, scored independently by 38 Dutch PRPP trained occupational therapists were included in the analysis. The random error between raters was analyzed with two-way random Intraclass Correlation Coefficients (ICC). (2) Four videos were scored twice by 30 raters (6 week time interval). The random error within raters was analyzed using one-way random ICC's. (3) Internal consistency study: data of 190 persons with PD were analyzed using Cronbach's alpha (α). RESULTS: Inter-rater reliability ranged from slight to moderate (ICC= 0.06-0.43). The mean intra-rater reliability ranged from moderate to almost perfect (ICC= 0.60-0.83). Internal consistency is good (α = 0.60-0.86). CONCLUSION: The limited inter-rater reliability but adequate intra-rater reliability and internal consistency show the feasibility of the PRPP when used for persons with PD. Implications for reliable clinical use are discussed.