Sonja Mensch

83 Reliability of Movakic Analyses For the assessment of test-retest reliability, baseline (T0) Movakic scores were compared with T1-scores at three months. In case an event was recorded that might have caused a change between T0 and T1, another 3-month period was selected in which no such events had occurred. For the assessment of inter-rater reliability, baselinemeasurements of two therapists were used. To test relative reliability, which is the degree in which children maintain their position in a sample over repeated measurements, two-way mixed model intraclass correlation coefficients (ICC) were calculated with 95%-confidence intervals between the Movakic scores of T0 en T1 (for test-retest reliability) and between the Movakic scores of the two raters (for inter-rater reliability). ICC’s for test-retest and inter-rater reliability were also calculated for each of the four groups of motor abilities. Reliability was classified as excellent (> 0.75), good (0.60–0.74), fair (0.40-0.59), or poor (<0.40) (Cicchetti & Sparrow., 1981). The ICC does not provide information about the degree in which actual scores for an individual vary over repeated measurements (absolute reliability). The smaller the differences, the higher the absolute reliability. This was done by performing Bland & Altman analyses (Bland & Altman., 1986). First, we calculated for each child the distance (absolute difference) between T0 and T1 Movakic scores (for test-retest reliability) and between both T0 Movakic scores per rater (for inter-rater reliability). Next, for test- retest reliability mean Movakic scores at T0 and T1 were plotted against the individual differences between T0 and T1, for inter-rater reliability mean Movakic scores were plotted against individual differences between both raters at T0. Adequate reliability is represented by small differences from the mean (within one standard deviation (SD) of the mean). All calculations were performed in IBM Statistical Package for the Social Sciences version 21. RESULTS Population Sixty children with a mean age of 7.7 years (range 2-16 years) were selected in 15 different day-care centers. Characteristics of the study population are listed in table 3.