Mediation results
| . | Mediator: Health Care Usage (2017) . | Mediator: Second Pregnancy . | Mediator: Contraceptive Use . | Mediator: Transport Use . | ||||
|---|---|---|---|---|---|---|---|---|
| . | (1) . | (2) . | (3) . | (4) . | (5) . | (6) . | (7) . | (8) . |
| A. HAZ | ||||||||
| ITT on mediator | 0.1020* | 0.1140* | −0.0376* | −0.0378† | 0.0653* | 0.0544† | 0.5630* | 0.5520* |
| (0.0442) | (0.0455) | (0.0188) | (0.0194) | (0.0311) | (0.0323) | (0.0854) | (0.0856) | |
| ACME | 0.0319† | 0.0385† | −0.056* | −0.0449† | 0.0197 | 0.0199† | −0.0380 | −0.0385 |
| [−0.0070, 0.0911] | [−0.0038, 0.0918] | [−0.1394, −0.0025] | [−0.1300, 0.0077] | [−0.0125, 0.0696] | [−0.0085, 0.0641] | [−0.1299, 0.0470] | [−0.1296, 0.0481] | |
| ρ, where ACME = 0 | .1 | .1 | .3 | .2 | — | .075 | — | — |
| Number of observations | 406 | 404 | 406 | 404 | 406 | 404 | 406 | 404 |
| B: CREDI z Score | ||||||||
| ITT on mediator | 0.1147* | 0.1017* | −0.0266 | −0.0284 | 0.0287 | 0.0338 | 0.5474* | 0.5470* |
| (0.0493) | (0.0508) | (0.0310) | (0.0322) | (0.0342) | (0.0355) | (0.1115) | (0.1106) | |
| ACME | 0.0030 | 0.0029 | 0.0012 | 0.0017 | 0.0008 | 0.0021 | −0.0001 | 0.0021 |
| [−0.0170, 0.0251] | [−0.0154, 0.0227] | [−0.0104, 0.0143] | [−0.0097, 0.0151] | [−0.0093, 0.0133] | [−0.0079, 0.0152] | [−0.0309, 0.0303] | [−0.0271, 0.0313] | |
| ρ, where ACME = 0 | — | — | — | — | — | — | — | — |
| Number of observations | 332 | 332 | 362 | 362 | 362 | 362 | 362 | 362 |
| Mother and Child Controls | X | X | X | X | ||||
| . | Mediator: Health Care Usage (2017) . | Mediator: Second Pregnancy . | Mediator: Contraceptive Use . | Mediator: Transport Use . | ||||
|---|---|---|---|---|---|---|---|---|
| . | (1) . | (2) . | (3) . | (4) . | (5) . | (6) . | (7) . | (8) . |
| A. HAZ | ||||||||
| ITT on mediator | 0.1020* | 0.1140* | −0.0376* | −0.0378† | 0.0653* | 0.0544† | 0.5630* | 0.5520* |
| (0.0442) | (0.0455) | (0.0188) | (0.0194) | (0.0311) | (0.0323) | (0.0854) | (0.0856) | |
| ACME | 0.0319† | 0.0385† | −0.056* | −0.0449† | 0.0197 | 0.0199† | −0.0380 | −0.0385 |
| [−0.0070, 0.0911] | [−0.0038, 0.0918] | [−0.1394, −0.0025] | [−0.1300, 0.0077] | [−0.0125, 0.0696] | [−0.0085, 0.0641] | [−0.1299, 0.0470] | [−0.1296, 0.0481] | |
| ρ, where ACME = 0 | .1 | .1 | .3 | .2 | — | .075 | — | — |
| Number of observations | 406 | 404 | 406 | 404 | 406 | 404 | 406 | 404 |
| B: CREDI z Score | ||||||||
| ITT on mediator | 0.1147* | 0.1017* | −0.0266 | −0.0284 | 0.0287 | 0.0338 | 0.5474* | 0.5470* |
| (0.0493) | (0.0508) | (0.0310) | (0.0322) | (0.0342) | (0.0355) | (0.1115) | (0.1106) | |
| ACME | 0.0030 | 0.0029 | 0.0012 | 0.0017 | 0.0008 | 0.0021 | −0.0001 | 0.0021 |
| [−0.0170, 0.0251] | [−0.0154, 0.0227] | [−0.0104, 0.0143] | [−0.0097, 0.0151] | [−0.0093, 0.0133] | [−0.0079, 0.0152] | [−0.0309, 0.0303] | [−0.0271, 0.0313] | |
| ρ, where ACME = 0 | — | — | — | — | — | — | — | — |
| Number of observations | 332 | 332 | 362 | 362 | 362 | 362 | 362 | 362 |
| Mother and Child Controls | X | X | X | X | ||||
Notes: For all columns, the unit of observation is the child. Results presented are for a causal mediation analysis of the ITT results on HAZ and CREDI z scores. Maternal and child controls include woman-level controls, such as the total number of children alive, educational attainment (primary or less vs. secondary and higher), age, religion (Christian vs. other), tribal ethnicity (Chewa vs. other), and age at sexual debut; they also include child-level controls, such as biological sex, as well as birth order, birth month, and a spline fitted to the child's age with knots at 6, 12, 18, 24, and 30 months. These same controls adjust ITT estimations on the mediator. Heteroskedasticity-robust standard errors, clustered by mother, are shown in parentheses; bootstrapped confidence intervals are included in brackets.
†p < .10; *p < .05