Skip to main content
. Author manuscript; available in PMC: 2015 Jan 22.
Published in final edited form as: J Hum Resour. 2013 Mar 1;48(4):945–968. doi: 10.3368/jhr.48.4.945

Table 4.

Treatment effects on IQ z-score by low-income status using IHDP HLBW sample with ECLS-B weights.

Outcome (sample size) Model
A B C
Age 1 IQ (n=330) Treatment 0.109 (0.132) 0.112 (0.133) 0.065 (0.177)
Low income −0.037 (0.122) −0.072 (0.171)
Treatment x (low income) 0.097 (0.253)
Age 2 IQ (n=322) Treatment 0.793*** (0.160) 0.878*** (0.223) 0.433* (0.219)
Low income −0.875*** (0.244) −1.181*** (0.270)
Treatment x (low income) 0.872** (0.280)
Age 3 IQ (n=328) Treatment 0.903*** (0.147) 1.001*** (0.181) 0.323 (0.210)
Low income −1.017*** (0.192) −1.482*** (0.240)
Treatment x (low income) 1.319*** (0.308)
Age 5 IQ (n=295) Treatment 0.102 (0.116) 0.148 (0.166) −0.264 (0.201)
Low income −0.509* (0.246) −0.820*** (0.231)
Treatment x (low income) 0.861*** (0.201)
Age 8 IQ (n=311) Treatment 0.156 (0.158) 0.224 (0.169) −0.067 (0.323)
Low income −0.595** (0.185) −0.806*** (0.196)
Treatment x (low income) 0.572 (0.361)

Coefficient significance (within site correlation corrected standard errors):

*

0.10

**

0.05

***

0.01. All models also condition on child gender, birth weight, gestational age at birth, neonatal health index, and site indicators. Estimates in appendix.