Page 26 - IJPS-4-1
P. 26

Transition to first marriage in China


           Table 4. Joint distribution (in percentage) of husband’s and wives’ education, overall and by marriage cohort (n=19,340).
           Husband’s education                                              Wife’s education
                                    Primary school and below  Junior middle   High school  College and more  Total
           All respondents
             Primary school and below      22.6               4.7          0.8            0.1        28.1
             Junior middle                 14.4               16.8         3.9            0.7        35.7
             High school                    3.8               6.9          8.2            2.0        20.9
             College and more               0.5               1.7          4.0            9.0        15.2
             Total                         41.4               30.1         16.9          11.7        100.0
           1966–1979 (n=4,080)
             Primary school and below      46.1               4.2          0.6            0.0         51
             Junior middle                 17.9               9.1          2.2            0.6        29.9
             High school                    5.4               4.0          2.9            0.7        13.0
             College and more               1.1               1.5          2.1            1.4         6.2
             Total                         70.5               18.8         7.8            2.8        100.0
            1980–1991 (n=7,437)
             Primary school and below      21.0               5.2          1.0            0.0        27.2
             Junior middle                 16.3               16.9         4.9            0.4        38.5
             High school                    5.0               8.0          10.1           1.4        24.5
             College and more               0.5               1.7          3.6            3.9         9.7
             Total                         42.8               31.9         19.6           5.7        100.0
            1992–2000 (n=4,634)
             Primary school and below      16.2               5.0          0.7            0.0        21.9
             Junior middle                 14.1               21.6         3.7            0.9        40.3
             High school                    2.0               7.4          7.9            2.6        19.9
             College and more               0.3               1.7          4.8           10.1        17.9
             Total                         32.7               35.8         17.1          14.4        100.0
           2001–2015 (n=3,189)
             Primary school and below       5.7               3.5          0.7            0.2        10.0
             Junior middle                  5.8               19.0         4.2            1.1        30.0
             High school                    1.6               7.4          10.9           4.3        24.2
             College and more               0.2               2.0          5.9           27.6         35.8
            Total                          13.3               31.9         21.7          33.2        100.0


           Table 5. Goodness-of-fit results of log linear models of assortative mating (n=20,333).
           Model                                    DF             L 2             BIC              Δ
           1. Conditional independence model.       35           10543.5          10196.3          27.6
           2. Multiplicative layer, quasi-indep.    28            2397.8          2120.1           9.1
           3. Multiplicative layer, uniform assoc.  31            526.0            218.5           6.0
           4. Multiplicative layer, R-C II          29            468.6            180.9           5.5
           5. Multiplicative layer, quasi-symmetry  26             96.0           −161.9           1.9
           6. Multiplicative layer, full interaction  23           81.3           −146.9           1.8
           Note: Model terms (number of parameters): DF=Degrees of freedom; L =The log likelihood ratio Chi-square statistic; BIC=L -(DF) ln (n); Δ represents the dissimilarity index
                                                  2
                                                                                2
           between observed and predicted frequencies (in percentage). All models are estimated using the Stata 14 software package



           20                                              International Journal of Population Studies | 2018, Volume 4, Issue 1
   21   22   23   24   25   26   27   28   29   30   31