Table 4. Gender interaction for learning outcomes.
numeracy | literacy | total | ||||
---|---|---|---|---|---|---|
(1) | (2) | (3) | (4) | (5) | (6) | |
Panel A: treatment effects by gender | ||||||
treat × female | 0.196* (0.0974) |
0.149* (0.0798) |
0.215** (0.104) |
0.148* (0.0792) |
0.217** (0.0999) |
0.160** (0.0777) |
treat | -0.0429 (0.0871) |
-0.00411 (0.0858) |
-0.0678 (0.120) |
-0.0209 (0.108) |
-0.0562 (0.0946) |
-0.0104 (0.0866) |
female | -0.112 (0.0813) |
-0.561 (0.444) |
-0.0250 (0.0948) |
0.136 (0.168) |
-0.0801 (0.0881) |
-0.304 (0.231) |
controls included? | no | yes | no | yes | no | yes |
N | 2052 | 2052 | 2053 | 2053 | 2031 | 2031 |
R2 | 0.00405 | 0.232 | 0.00481 | 0.246 | 0.00447 | 0.274 |
Panel B: female only | ||||||
treat | 0.154 (0.0995) |
0.145 (0.0954) |
0.147 (0.130) |
0.127 (0.117) |
0.160 (0.114) |
0.149 (0.105) |
controls included? | no | yes | no | yes | no | yes |
N | 1010 | 1010 | 1012 | 1012 | 998 | 998 |
R2 | 0.00752 | 0.236 | 0.00547 | 0.261 | 0.00788 | 0.284 |
In Panel A, “treat × female” represents the differential effect of treatment for female respondents compared to male respondents. The uninteracted treatment coefficient presents treatment effects for male respondents. Controls (in columns 2, 4, and 6) include gender, age, and survey round, and all interactions of these variables. Robust standard errors clustered at parish level in parentheses.
* p< .1,
** p< .05,
*** p< .01