Table 4.
Average effect of heavy drinking on postsecondary degree receipt, by gender and socioeconomic background based on one-to-one nearest-neighbor matching without replacement
| Variable | Sample size, n |
||||||
| Heavy drinkersa | Moderate and nondrinkers | Difference | SE | t | Treated | Total | |
| By gender | |||||||
| Females | |||||||
| Unmatched | .292 | .309 | -.017 | .021 | -0.78 | ||
| Matched | .292 | .318 | -.025 | .028 | -0.91 | 554 | 3,804 |
| Males | |||||||
| Unmatched | .293 | .339 | -.046 | .017 | -2.67* | ||
| Matched | .293 | .365 | -.072 | .021 | -3.43* | 1,000 | 3,727 |
| By gender and socioeconomic background | |||||||
| Females | |||||||
| Working class | |||||||
| Unmatched | .219 | .251 | -.032 | .023 | -1.39 | ||
| Matched | .219 | .239 | -.020 | .030 | -0.67 | 402 | 2,996 |
| Nonmanual | |||||||
| Unmatched | .487 | .538 | -.051 | .045 | -1.14 | ||
| Matched | .487 | .500 | -.013 | .058 | -0.23 | 152 | 808 |
| Males | |||||||
| Working class | |||||||
| Unmatched | .236 | .275 | -.039 | .018 | -2.1* | ||
| Matched | .236 | .317 | -.080 | .023 | -3.57* | 783 | 2,911 |
| Nonmanual | |||||||
| Unmatched | .498 | .568 | -.070 | .039 | -1.77 | ||
| Matched | .498 | .553 | -.055 | .048 | -1.15 | 217 | 816 |
a
Heavy drinking for females is defined as four more or units in the past week, whereas heavy drinking for males is defined as five or more units. We used the “psmatch2” command in Stata 10.0 to estimate the matching procedure.
*
p < .05 (two-tailed test).