Identifying the Smoking and Smokeless Tobacco-Related Predictors on Frequencies of Heavy Vehicle Traffic Accidents in Bangladesh: Linear and Binary Logistic Regression-Based Approach

Table 9

Regression equation, coefficients, model summary, goodness-of-fit tests of the regression, and fits and diagnostics for unusual observations.

SmokeorTobaccoStatus			SmokeDurDrive
Regression equation	NO			NO
	NO			YES
	YES			NO
	YES			YES
Coefficients	Term			Coef	SE coef	VIF
	Constant			0.738	0.212
	CigUseYr			0.0531	0.0161	1.63
	SmokeorTobaccoStatus
	YES			−1.183	0.318	1.70
	SmokeDurDrive
	YES			0.684	0.245	1.28
Model summary	Deviance			Deviance
	R-sq			R-sq (adj)	AIC	AICc	BIC
	4.69%			4.15%	533.12	533.22	549.32
Goodness-of-fit tests of the regression	Test			DF	Chi-square	value
	Deviance			420	525.12	≤0.001
	Pearson			420	429.22	0.367
	Hosmer–Lemeshow			7	13.31	0.065
Fits and diagnostics for unusual observations	Obs	Observed probability	Fit	Residuals	Std residuals	Type of residuals
	49	≤0.001	0.8619	−1.9900	−2.00	R
	205	1.0000	0.8427	0.5851	0.60		X
	250	1.0000	0.8562	0.5572	0.57		X
	274	≤0.001	0.8957	−2.1261	−2.14	R
	341	≤0.001	0.8204	−1.8531	−1.88		X

R = large std residuals; X = unusual std residuals.