Measuring the health of the Indian elderly: evidence from National Sample Survey data

Chen, Bradley; Mahal, Ajay

doi:10.1186/1478-7954-8-30

Population Health Metrics

Table 5 Test for differential response by education

From: Measuring the health of the Indian elderly: evidence from National Sample Survey data

	Rural				Urban
	North	East	South	West	North	East	South	West
Males
Log likelihood: <primary	-2459.77	-1874.81	-947.86	-598.54	-561.68	-377.78	-394.38	-228.74
Log likelihood: primary+	-391.29	-471.61	-387.01	-125.45	-762.63	-600.70	-661.10	-586.81
Log likelihood sum (unrestricted model, L ^U)	-2851.06	-2346.41	-1334.88	-723.99	-1324.31	-978.48	-1055.47	-815.55
Log likelihood restricted model (L ^R)	-2856.40	-2350.25	-1338.57	-729.46	-1325.46	-986.37	-1060.74	-817.75
X ² -first stage test statistic: -2*( L ^R-L ^U)	10.67	7.68	7.39	10.94	2.29	15.77	10.53	4.41
Sample size	3793	2820	1832	1032	1820	1199	1499	1054
P- value (8 degrees of freedom) ¹	0.22	0.47	0.50	0.21	0.97	0.05	0.23	0.82
Log likelihood partially-restricted model (L ^Rα)	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.
X ²-test statistic for cut-point shift: -2*( L ^Rα-L ^U)	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.
P- value (2 degrees of freedom) ²	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.
X ²-test statistic for g (.) function: -2*( L ^R-L ^Rα)	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.
P- value (6 degrees of freedom) ³	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.
Females
Log likelihood: <primary	-2810.45	-2102.21	-1320.54	-801.73	-1183.61	-801.69	-876.94	-551.19
Log likelihood: primary+	-79.89	-56.35	-169.58	-18.70	-307.07	-291.29	-329.00	-330.72
Log likelihood sum (unrestricted model, L ^U)	-2890.34	-2158.56	-1490.12	-820.43	-1490.68	-1092.98	-1205.94	-881.91
Log likelihood restricted model (L ^R)	-2893.46	-2175.41	-1495.02	-825.57	-1494.84	-1098.79	-1213.65	-884.54
X ² -first stage test statistic: -2*( L ^R-L ^U)	6.23	33.69	9.81	10.27	8.32	11.63	15.42	5.24
Sample size	4011	2692	2017	1221	2057	1306	1841	1265
P- value (8 degrees of freedom) ¹	0.62	4.6 × 10-5	0.28	0.25	0.40	0.17	0.05	0.73
Log likelihood partially-restricted model (L ^Rα)	n.a.	-2167.19	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.
X ²-test statistic for cut-point shift: -2*( L ^Rα-L ^U)	n.a.	17.26	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.
P- value (2 degrees of freedom) ²	n.a.	0.0002	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.
X ²-test statistic for g (.) function: -2*( L ^R-L ^Rα)	n.a.	16.43	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.
P- value (6 degrees of freedom) ³	n.a.	0.012	n.a.	n.a.	n.a.	n.a.	n.a.	n.a.

¹6 parameters for H^o (α) and 3 cut-points (τ) were estimated the unrestricted model for two subgroups (9 × 2 = 18); 6 parameters for H^o (α), 1 parameters for membership (θ) and 3 cut-points (τ) were estimated for the restricted model. The degree of freedom is equal to the difference in the number of estimated parameters between these two models (18-10 = 8).
² 16 parameters, including 6 parameters for H^o (α) for the two subgroups (6 × 2 = 12), 1 parameters for membership (θ) and 3 cut-points (τ) were estimated for the partially-restricted model. The degree of freedom is equal to the difference of number of estimated parameters between unrestricted and partially-restricted models (18-16 = 2).
³ The degree of freedom is equal to the difference of number of estimated parameters between partially-restricted models and restricted models (16-10 = 6).

Back to article page

ISSN: 1478-7954

Contact us

General enquiries: journalsubmissions@springernature.com