Skip to main content

Table 5 Characteristics of the secondary variables for the 191 samples composing the main dataset (obtained from first-order variables, Table 1)

From: Estimating upper-extremity function from kinematics in stroke patients following goal-oriented computer-based training

Variables Range [min, max] Mean SD r(FM-UE) r(CAHAI) r(BI)
15. Chronic Yes/no 57/134 − 0.25 − 0.17 0.14
   \({\texttt {<}}\)Subacute\({\texttt {>}}\) Yes/no 74/117
16. Acute Yes/no 60/131 0.32 0.18 − 0.15
17. Diff. work area (\(m^2\)) [− 1.3, 1.6] 0.25 0.50 − 0.31 − 0.29 (− 0.069)
18. Diff. distance covered (m) [− 160, 120] 11 34 − 0.32 − 0.26 − 0.12
19. Diff. performance (% success) [− 0.22, 0.53] 0.087 0.13 − 0.26 − 0.22 − 0.16
20. Diff. maximum reaching speed (m/s) [− 61, 98] 8.1 23 − 0.22 − 0.22 − 0.15
21. Diff. difficulty level reached [− 0.47,0.63] 0.11 0.19 − 0.22 − 0.15 (− 0.058)
22. Diff. smoothness (mm) [− 2.6, 4.4] 0.34 0.75 − 0.29 − 0.28 − 0.20
23. Diff. TGDM (m) [− 0.035, 0.084] 0.015 0.021 − 0.52 − 0.47 − 0.24
24. Log. time since stroke (days) [1.6, 8.0] 4.8 1.7 − 0.24 (− 0.076) 0.36
25. Log. sessions completed so far [0.0, 3.9] 1.8 0.9 0.29 0.38 0.42
26. Log. work area [− 4.3, 0.62] − 1.3 0.9 0.40 0.38 0.19
27. Log. distance covered [0.92, 5.5] 3.8 0.7 0.26 0.32 0.34
28. Log. maximum reaching speed [1.0, 4.5] 2.6 0.73 0.26 0.20 (0.045)
29. Log. difficulty level reached [− 0.16, 0.60] 0.25 0.14 0.44 0.50 0.45
30. Log. smoothness [− 1.7, 1.4] 0.012 0.48 0.48 0.46 0.33
31. Log. TGDM [− 4.7, − 2.1] − 2.9 0.48 0.52 0.56 0.43
  1. The r columns refer to the Pearson correlation coefficients with the FM-UE, CAHAI, and BI clinical scales, respectively. Correlations below the threshold \(r \sim 0.081\) (Fig. 12 in Appendix) are in parenthesis. From the time since stroke, we obtain the categories Acute (5–90 days), Sub-acute (3–12 months), and Chronic (over 1 year). The variables obtained directly from the RGS system log files are in Italic type. The Diff. variables are obtained as the difference between the value observed for the less affected arm and the value for the more affected one. The Log. variables are obtained as the natural logarithm of the corresponding first-order variables