Table 3.
Robustness Checks: Dealing with Measurement Error in Perceived Depression
| Model | (1) Probit | (2) Probit | (3) Probit | (4) Probit | (5) Probit | (6) Probit |
|---|---|---|---|---|---|---|
| Variable | Whether depressed | Whether depressed | Whether depressed | Whether depressed | Whether depressed | Whether depressed |
| BMI | −0.116*** (0.013) | −0.125*** (0.013) | −0.113*** (0.013) | −0.117*** (0.014) | −0.113*** (0.014) | −0.097*** (0.014) |
| BMI_squared | 0.002*** (0.000) | 0.002*** (0.000) | 0.002*** (0.000) | 0.002*** (0.000) | 0.002*** (0.000) | 0.002*** (0.000) |
| Demographic characteristics | No | Yes | Yes | Yes | Yes | Yes |
| Human capital characteristics | No | No | Yes | Yes | Yes | Yes |
| Social characteristics | No | No | Yes | Yes | Yes | Yes |
| Working characteristics | No | No | No | Yes | Yes | Yes |
| Family characteristics | No | No | No | No | Yes | Yes |
| Year dummies | No | No | No | No | No | Yes |
| Province dummies | No | No | No | No | No | Yes |
| Constant | 1.042*** (0.160) | 0.445*** (0.165) | 0.779*** (0.173) | 0.804*** (0.185) | 0.628*** (0.190) | 0.043 (0.199) |
| Observations | 25,197 | 25,197 | 25,036 | 23,691 | 23,515 | 23,515 |
| Pseudo R2 | 0.003 | 0.009 | 0.023 | 0.026 | 0.031 | 0.043 |
Notes: ***Indicates significance at the level of 1%. The values in parentheses are standard errors robust to heteroskedasticity. Yes means the corresponding variables are controlled in the regression, while No means not controlled. Pseudo R2 is reported for nonlinear models to assess goodness of fit.