From capital to habitus: class differentiation of family educational patterns in urban China

The Journal of Chinese Sociology

Table 3 Regression results of factors influencing cultural capital investment

Variables	Model 1 (NBR)	Model 2 (OLS)	Model 3 (logistic)
Variables	Cultural activities	Books in the home	Shadow education
	B (S.E.)	B (S.E.)	B (S.E.)
Constant	−0.341 (0.166)*	−37.62 (12.42)*	−3.961 (0.504)***
Eighth grader^a	−0.256 (0.049)***	13.598 (4.174)**	−0.659 (0.144)***
Male student^b	−0.054 (0.018)**	−0.575 (2.638)	−0.234 (0.062)***
Han student^c	0.097 (0.065)	6.654 (5.351)	0.355 (0.186)
Father interviewed^d	−0.048 (0.021)*	−3.574 (2.796)	0.353 (0.079)***
Urban parent^e	0.181 (0.034)***	19.829 (3.313)***	0.772 (0.097)***
Parental higher edu.^f	0.055 (0.028)*	57.332 (4.470)***	0.415 (0.115)***
Ln of per family income	0.162 (0.021)***	8.638 (1.514)***	0.504 (0.065)***
Parents’ occu. class^g
Upper middle class	0.126 (0.029)***	24.620 (4.757)***	0.423 (0.117)***
Lower middle class	0.084 (0.028)**	7.979 (3.281)*	0.368 (0.079)***
Parents’ subjective class^h
Upper middle class	0.168 (0.038)***	23.416 (3.919)***	0.154 (0.125)
Lower middle class	0.076 (0.039)*	4.320 (4.340)	0.176 (0.126)
Unable to classify	0.109 (0.037)**	8.138 (4.224)	−0.015 (0.118)

Table 3 includes negative binomial regression of participation in cultural activities, OLS regression of number of books in the home, and logistic regression of attendance at shadow education. N = 5411. The reference categories for each variable are as follows: ^afourth grader, ^bfemale, ^cminority, ^dmother, ^erural households, ^fdid not receive higher education, ^goccupational lower class, and ^hsubjective lower class. Coefficients in the table are nonstandardized regression coefficients; standard errors in parentheses are robust standard errors controlling for classroom cluster effects
*p < 0.05, **p < 0.01, ***p < 0.001