Table 8. The treatment effect of normative policies on Culturees.
| Cutoff at week 0 | Cutoff at week 5 | |||||||
|---|---|---|---|---|---|---|---|---|
| (1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | |
| MSE-bandwidth estimate | -0.246*** (0.002) |
-0.200*** (0.032) |
-0.270** (0.049) |
-0.229*** (0.038) |
0.268*** (0.038) |
0.281*** (0.061) |
0.287*** (0.045) |
0.283*** (0.064) |
| Bandwidth | 2.74 | 14.73 | 10.00 | 15.79 | 3.07 | 9.83 | 10.00 | 7.17 |
| Double MSE-bandwidth estimate | -0.232*** (0.033) |
-0.207*** (0.029) |
-0.264*** (0.035) |
-0.201*** (0.033) |
0.317*** (0.059) |
0.265*** (0.046) |
0.222*** (0.043) |
0.274*** (0.051) |
| Bandwidth | 5.48 | 29.46 | 20.000 | 31.58 | 6.14 | 19.66 | 20.00 | 14.34 |
| Triple MSE-bandwidth estimate | -0.206*** (0.025) |
-0.204*** (0.031) |
-0.303*** (0.034) |
-0.201*** (0.033) |
0.315*** (0.048) |
0.255*** (0.044) |
0.222*** (0.043) |
0.245*** (0.044) |
| Bandwidth | 8.22 | 44.19 | 30.000 | 47.37 | 9.21 | 29.49 | 30.00 | 21.51 |
| N | 50 | 50 | 50 | 50 | 50 | 50 | 50 | 50 |
| Number of texts | 2000 | 2000 | 2000 | 2000 | 2000 | 2000 | 2000 | 2000 |
| Polynomial | Linear | Quadratic | Linear | Quadratic | Linear | Quadratic | Linear | Quadratic |
| Kernel function | triangular | triangular | uniform | uniform | triangular | triangular | uniform | uniform |
Standard errors clustered by Culturees. Asterisks denote significance levels (* = 0.10, ** = 0.05, *** = 0.01). Each cell presents a separate estimate from a rdms analysis. The results of Columns (3) and (7) derive from a customized bandwidth as a result of the failure of MSE-based local linear polynomial along with the kernel function (uniform) estimate.