1. | The change scores (differences between baseline and 6-month assessments) of a physical capacity task will be able to distinguish between patients with and without meaningful improvementa as classified by a construct-specific GPE scale (area under the ROC curve ≥0.70) [24, 37, 56].b |
2. | The change scores of a physical capacity task will yield greater misclassifications of improved and unchanged patients on a ROC curve when that classification is based on a generic GPE scale rather than construct-specific GPE scales [24, 37, 56].b |
3. | The change scores of the four physical capacity tasks will be correlated ≥0.50 to each other in the expected direction [17, 22].c |
4. | The correlations between change scores of physical capacity tasks and the ODI will be at least 0.10 weaker than the correlations between the change scores among the physical capacity tasks themselves [13, 20]. |
5. | The correlations between change scores of a physical capacity task and VAS on back pain intensity will be at least 0.10 weaker than the correlations between change scores of the physical capacity task and the ODI [13, 17]. |