Table 7. Goodness-of-fit and classification accuracy of the binary logistic predictive models predicting adult performance level (professional vs. non-professional; N = 121).
Type of correction used in predictive motorperformance model |
Omnibus test of model coefficients |
Hosmer-Lemeshow test |
Model fit |
Correct classification |
>2(8) |
p |
>2(8) |
p |
R2N |
% |
None (raw scores) |
23.53 |
.003 |
6.02 |
.645 |
0.24 |
64.5 |
Mirwald |
21.40 |
.006 |
4.84 |
.775 |
0.22 |
66.9 |
Moore-1 |
21.39 |
.006 |
12.11 |
.146 |
0.22 |
67.8 |
Moore-2 |
23.02 |
.003 |
6.12 |
.634 |
0.24 |
64.5 |
Fransen-1 |
21.53 |
.006 |
2.45 |
.964 |
0.22 |
65.3 |
Fransen-2 |
21.51 |
.006 |
4.24 |
.835 |
0.22 |
66.1 |
%PAH |
18.40 |
.018 |
8.79 |
.361 |
0.19 |
68.6 |
%RAH |
28.23 |
<.001 |
3.79 |
.876 |
0.28 |
66.1 |
%PAH = attained percentage of the predicted adult height, %RAH = attained percentage of the real adult height.