Table 3 Results of binary logit and...

Table 3

Results of binary logit and probit models

Factors	Binary logit model			Probit model
Factors	B	S.E.	Sig	B	S.E.	Sig
Age	−0.386	0.316	0.222	−0.176	0.155	0.255
Gender	1.530	1.314	0.245	0.748	0.649	0.249
Level of education	0.147	0.159	0.355	0.069	0.080	0.389
Employment status	−0.940	0.586	0.109	−0.447	0.276	0.104
Income per month	1.228*	0.637	0.054	0.646**	0.329	0.049
Income instability	4.660**	1.839	0.011	2.407**	0.976	0.014
Shrinking work opportunity	2.992***	0.857	0.000	1.511**	0.439	0.000
Number of dependents in family	1.699**	0.741	0.022	0.766**	0.373	0.039
Other’s financial contribution in family	−1.782**	0.862	0.039	−0.835**	0.421	0.048
savings	−1.818**	0.843	0.031	−0.920**	0.421	0.029
Comparison of household expense to pre-pandemic expense	−1.680***	0.574	0.003	−0.876**	0.296	0.003
major sources of expenditure	1.359	0.840	0.106	0.656	0.439	0.135
Constant	−6.220	5.049	0.218	−3.160	2.622	0.228

Factors	Binary logit model			Probit model
Factors	B	S.E.	Sig	B	S.E.	Sig
Age	−0.386	0.316	0.222	−0.176	0.155	0.255
Gender	1.530	1.314	0.245	0.748	0.649	0.249
Level of education	0.147	0.159	0.355	0.069	0.080	0.389
Employment status	−0.940	0.586	0.109	−0.447	0.276	0.104
Income per month	1.228*	0.637	0.054	0.646**	0.329	0.049
Income instability	4.660**	1.839	0.011	2.407**	0.976	0.014
Shrinking work opportunity	2.992***	0.857	0.000	1.511**	0.439	0.000
Number of dependents in family	1.699**	0.741	0.022	0.766**	0.373	0.039
Other’s financial contribution in family	−1.782**	0.862	0.039	−0.835**	0.421	0.048
savings	−1.818**	0.843	0.031	−0.920**	0.421	0.029
Comparison of household expense to pre-pandemic expense	−1.680***	0.574	0.003	−0.876**	0.296	0.003
major sources of expenditure	1.359	0.840	0.106	0.656	0.439	0.135
Constant	−6.220	5.049	0.218	−3.160	2.622	0.228

Note(s): *, ** and *** denote degree of significance at 10%, 5% and 1% level, respectively. Number of observations = 434. For binary model: Hosmer and Lemeshow Test: Chi-square = 1.003**; −2 Log likelihood = 65.946; Cox and Snell R Square = 0.126; Nagelkerke R Square = 0.501 and for probit model: McFadden R-square = 0.470; Log likelihood = −32.637; Restr. log-likelihood = −61.616; overall accuracy (correctly predicted) = 96.5%

Source(s): Authors’ own work

[ViewLarge]

Sharing Unavailable