2.4. Statistical Analyses

Normality of distribution was tested with the Kolmogorov–Smirnov test.

To depict the relationship between gross motor capacity and physical daily life performance, we correlated GMFM-66 scores with PAS and pattern complexity using Pearson correlations in case of normal or Spearman’s rho in case of non-normal distribution of data. Correlations were interpreted as follows: 0.00–0.25 no to little; 0.25–0.50 fair; 0.50–0.75 moderate to good; 0.75–1.00 very good to excellent agreement [36].

To investigate the ability of the GMFM-66 score to discriminate between achievers and non-achievers for each PAS, we performed receiver operating characteristics (ROC) analysis. The highest Youden Index (= sensitivity + specificity −1) was extracted to depict the cut-off values with the best proportion between sensitivity and specificity. Scatter plots were used to visualize ROC analyses with respective cut-off values. Cut-off values and scatter plots were only computed for PAS where ROC resulted in an area under the curve (AUC) significantly different from 0.5 (which represents chance).

Retrospective analyses (Spearman correlation, alpha = 0.05) showed age to be associated with PAS 3, Hn, and LZV. Therefore, authors performed partial correlations correcting for age for those three variables.

Statistical analyses were performed using SPSS 26 (SPSS Inc., Chicago, IL, USA) and pairwise deletion was used for missing items. Alpha was set at 0.05, and the Holm method [37] was applied for multiple comparisons. The method was applied separately for comparisons of the GMFM-66 with PAS (7 comparisons) and with pattern complexity (4 comparisons), as these were treated and interpreted as two different constructs.