. 2019 Jul 17;14(7):e0219798. doi: 10.1371/journal.pone.0219798

Table 3. Statistic of predictive power and goodness-of-fit test in step-by-step regression analysis.

	Determination coefficient (R²_N)			Goodness-of-fit
Order in analysis	Cumulat.	Single	% relative	H-L test	% cases	Pearson
Physical activity	0.040	0.040	10.0	0.353	46.4	0.030
Socio-demographic	0.203	0.163	40.9	0.149	66.7	0.479
Age	0.177	0.137	34.3	0.429	54.0	-
Sex	0.182	0.005	1.3	0.065	68.2	-
Education	0.195	0.013	3.3	0.061	68.0	-
Housekeeper	0.203	0.008	2.0	0.149	66.7	-
Lifestyle	0.215	0.012	3.0	0.481	68.0	0.615
TV watching	0.208	0.005	1.3	0.356	67.4	-
Sleep	0.211	0.003	0.8	0.284	67.5	-
Nap	0.211	0.000	0.0	0.211	67.7	-
Smoker	0.215	0.003	0.8	0.391	68.0	-
Alcohol	0.215	0.000	0.0	0.481	68.0	-
Family history	0.221	0.006	1.5	0.167	67.9	0.530
BMI	0.399	0.178	45.1	0.141	74.9	0.968
Total	0.399	0.399	100	0.141	74.9	0.968

R²_N = Nagelkerke pseudo R²; H-L test = p-value of Hosmer and Lemeshow test; % cases = percentage of correctly predicted cases; Pearson = p-value of Pearson χ² test.