Skip to main content
. 2020 Aug 20;135(2):619–629. doi: 10.1007/s00414-020-02395-w

Table 2.

Summary statistics of the difference and the absolute value of the difference between the estimated age and the chronological age in JRA-affected subjects

Panel A: age estimation error,  E
Variable n (%) Demirjian mean (s.d.) Willems mean (s.d.) p Value
Sex Female 79 (77) 0.54 (0.94) − 0.32 (1.03) < 0.01
Male 24 (23) 0.27 (0.74) − 0.17 (0.97) 0.03
p Value 0.14 0.51
Under treatment Yes 40 (39) 0.59 (0.93) − 0.35 (1.01) < 0.01
No 63 (61) 0.41 (0.88) − 0.25 (1.02) < 0.01
p Value 0.35 0.77
Panel B: absolute age estimation error, |E|
Variable Demirjian median (1st–3rd quartile) Willems median (1st–3rd quartile) p Value
Sex Female 0.81 (0.27–1.00) 0.84 (0.36–1.14) 0.72
Male 0.54 (0.14–0.71) 0.69 (0.28–0.70) 0.31
p Value 0.03 0.17
Under treatment Yes 0.86 (0.27–1.13) 0.86 (0.55–1.10) 0.72
No 0.67 (0.19–0.81) 0.76 (0.24–1.05) 0.30
p Value 0.06 0.16

The p value refers to the significance of a t statistics in panel A and to the significance of Wilcoxon statistics in Panel B