Indonesian Higher Education 129
TABLE 5.7
Probit Estimation with Selectivity CorrectionVariables Public Private NR Private Islam Private Christian
Coeff./(t-stat) Coeff./(t-stat) Coeff./(t-stat) Coeff./(t-stat)
FOREGONE –0.221 –0.143 –0.157 –0.419
(0.012) (0.013) (0.019) (0.040)
TSRATIO –0.206 0.106 0.815 –0.663
(0.177) (0.152) (0.231) (0.440)
MALE 0.113 0.0921 0.032 0.348
(0.0252) (0.027) (0.046) (0.076)
ISLAM –0.022 –0.197 –0.037 0.022
(0.045) (0.052) (0.232) (0.076)
LANGINDO 0.120 0.065 0.056 0.234
(0.031) (0.031) (0.065) (0.098)
URBAN –0.009 –0.023 0.135 –0.255
(0.029) (0.032) (0.053) (0.089)**
SIBLING –0.034 –0.039 –0.035 –0.069
(0.010) (0.010) (0.018) (0.040)
WORKSMA –0.026 –0.038 0.039 –0.104
(0.046) (0.041) (0.066) (0.100)
SMAFAIL –0.090 –0.036 –0.638 0.000
(0.175) (0.107) (0.195) (0.000)
ZNEMSMA 0.048 0.046 0.063 –0.002
(0.013) (0.015) (0.022) (0.038)
FATHSHHE 0.178 0.026 0.081 0.154
(0.033) (0.040) (0.070) (0.096)
MOTHSHHE 0.165 0.189 0.003 0.161
(0.041) (0.053) (0.098) (0.120)
CONSTANT 0.176 0.527 0.183 0.772
(0.075) (0.083)** (0.256) (0.219)
λ –0.113 0.127 0.040 –0.006
(0.045)** (0.039)*** (0.049) (0.043)
Adjusted R2 0.345 0.300 0.235 0.730
N 11095 1622 1303 181
Notes: Standard errors are in parenthesis and heteroscedasticity consistent. Significance at 10 per cent
level, Significance at 5 per cent level, and Significance at 1 per cent level. Dummies representing
the province where individual was lived were also included in the model.
ZNEMSMP, PRIVATENRJH, PRIVATEISJH, and PRIVATECPOJH.
Following Wooldridge (2001), I will exclude all variables in the selection
equation from the structural equation or the probit higher education
participation equation. The objective of this exclusion is to avoid the
estimation from becoming very imprecise.