Table 6 Estimation results of probit...

Table 6

Estimation results of probit models

	FULL		Model A		Model B		Model C		Model D		Model E
	Coef	p	Coef	p	Coef	p	Coef	p	Coef	p	Coef	p
North–Centre	1.21	***	0.38	**	0.38	**	0.39	**	0.38	**	0.38	**
Woman	−0.28	***	−0.45	***	−0.45	***	−0.46	***	−0.46	***	−0.6	***
Graduate	0.09	*	0.18	**	0.19	**	0.19	***	0.30	*	0.18	**
High school	0.06		0.10	*	0.10	*	0.11	*	0.10	*	0.10	*
Hard work	−0.14	***	−0.14	**	−0.14	**	−0.63	***	−0.14	**	−0.14	**
Front office	−0.17	***	−0.20	***	0.01		−0.21	***	−0.20	***	−0.20	***
Employment rate	−0.04	***	−0.02	**	−0.02	**	−0.02	**	−0.02	**	−0.02	**
First generation	−0.44	***	−0.79	***	−0.62	***	−1.01	***	−0.75	***	−0.89	***
Second generation	−0.28	***	−0.53	***	−0.49	***	−0.62	***	−0.51	***	−0.59	***
Front office*1st Gen					−0.39	**
Front office*2st Gen					−0.10
Hard Work*1st Gen							0.78	***
Hard work*2st Gen							0.31	*
Graduate*1st Gen									−0.19
Graduate*2st Gen									−0.08
Woman*1st Gen											0.2
Woman*2st Gen											0.12
Constant	0.63	***	0.92	***	0.82	**	1.06	***	0.9	***	0.99	***
Log likelihood	−5277.579		−2903.66		−2895.58		−2882.98		−2902.59		−2902.34
BIC	10655.15		5892.57		5893.46		5868.26		5907.49		5906.99
AIC	10575.16		5827.32		5815.16		5789.97		5829.19		5828.69
Correctly classified	93.14%		71.58%		71.34%		71.66%		71.68%		71.77%
Brier score	0.062		0.195		0.194		0.193		0.195		0.195

	FULL		Model A		Model B		Model C		Model D		Model E
	Coef	p	Coef	p	Coef	p	Coef	p	Coef	p	Coef	p
North–Centre	1.21	***	0.38	**	0.38	**	0.39	**	0.38	**	0.38	**
Woman	−0.28	***	−0.45	***	−0.45	***	−0.46	***	−0.46	***	−0.6	***
Graduate	0.09	*	0.18	**	0.19	**	0.19	***	0.30	*	0.18	**
High school	0.06		0.10	*	0.10	*	0.11	*	0.10	*	0.10	*
Hard work	−0.14	***	−0.14	**	−0.14	**	−0.63	***	−0.14	**	−0.14	**
Front office	−0.17	***	−0.20	***	0.01		−0.21	***	−0.20	***	−0.20	***
Employment rate	−0.04	***	−0.02	**	−0.02	**	−0.02	**	−0.02	**	−0.02	**
First generation	−0.44	***	−0.79	***	−0.62	***	−1.01	***	−0.75	***	−0.89	***
Second generation	−0.28	***	−0.53	***	−0.49	***	−0.62	***	−0.51	***	−0.59	***
Front office*1st Gen					−0.39	**
Front office*2st Gen					−0.10
Hard Work*1st Gen							0.78	***
Hard work*2st Gen							0.31	*
Graduate*1st Gen									−0.19
Graduate*2st Gen									−0.08
Woman*1st Gen											0.2
Woman*2st Gen											0.12
Constant	0.63	***	0.92	***	0.82	**	1.06	***	0.9	***	0.99	***
Log likelihood	−5277.579		−2903.66		−2895.58		−2882.98		−2902.59		−2902.34
BIC	10655.15		5892.57		5893.46		5868.26		5907.49		5906.99
AIC	10575.16		5827.32		5815.16		5789.97		5829.19		5828.69
Correctly classified	93.14%		71.58%		71.34%		71.66%		71.68%		71.77%
Brier score	0.062		0.195		0.194		0.193		0.195		0.195

Note(s): p-value = *p < 0.05; **p < 0.01; ***p < 0.001

Although the percentage of accurate forecasts is adequate, we show another statistic, the Brier score (Brier, 1950), which is specifically designed to evaluate probability forecasts. It is defined as $Q R = \frac{\sum_{n = 1}^{N} {(r - d_{n})}^{2}}{N}$ where r is the predicted probability that the event will occur on the n-th occasion; $d_{n}$ takes value 1 if the event occurs on the n-th occasion and 0 otherwise. The Brier score ranges between 0 (if the predictions are always accurate) and 1 (if the predictions are always wrong)

[ViewLarge]