Table 6.

Model coefficient, p-value, goodness-of-fit and effect size of binary logistic regression mode, considering “why the battery is not recycled” and independent variables

	BLR model coefficient	BLR p-value	Hosmer and Lemeshow test (goodness-of-fit) for BLR, p value	Effect size measure
Recycling points are far from my household
Gender			0.841*	r = 0.140 r^2 = 0.0196
Age	0.025	0.848
Education
Household size	0.25	0.489
Income	0.285	0.012
I do not think battery can be recycled
Gender	0.126	0.783	0.298	r = 0.242 r^2 = 0.058564
Age	0.159	0.411
Education	− 0.703	0.052
Household size	0.89	0.016
Income	− 0.302	0.125
Do not know how and where to recycle
Gender	0.003	0.989	0.121	r = 0.288 r^2 = 0.082944
Age	− 0.055	0.55
Education	1.092	0
Household size	− 0.251	0.146
Income	− 0.236	0.005
I do not have time disposing of them
Gender			0.069*	r = 0.083 r^2 = 0.006889
Age
Education
Household size
Income	0.153	0.099

Bold numbers refer to significant correlations between the variables (p-value ≤ 0.05). Hosmer and Lemeshow test is for goodness-of-fit for the models. If the p value of the test is higher than 0.05, then the model is fitted. *Results of the revised model