Table 3 Nonparametric fuzzy regression discontinuity (RD) estimates—gender heterogeneity

MSE Optimal

 Bandwidth

3.2

3.6

2.9

3.8

2.6

3.9

 Conventional Point Estimate

0.3551

-0.3737

0.2127

-0.4074

-0.0588

0.5449

 Conventional 95% CI

[-0.2034, 0.9137]

[-1.406, 0.6585]

[-0.1408, 0.5663]

[-1.2016, 0.3867]

[-7.3082, 7.1906]

[-13.4954, 14.5853]

 Robust Point Estimate

0.4246

-0.5186

0.2917

-0.3544

-0.8639

1.0986

 Robust 95% CI

[-0.2357, 1.0849]

[-1.7377, 0.7004]

[-0.1159, 0.6993]

[-1.3216, 0.6128]

[-9.2719, 7.5442]

[-15.8644, 18.0615]

 Effective obs L|R

545|528

449|458

487|470

476|475

456|426

484|491

CER Optimal

 Bandwidth

2.1

2.4

1.9

2.6

1.8

2.7

 Conventional Point Estimate

0.5860

-0.6615

0.2209

-0.5006

-1.5189

4.0497

 Conventional 95% CI

[-0.1914, 1.3634]

[-2.2352, 0.9121]

[-0.1498, 0.5916]

[-1.5536, 0.5524]

[-9.9193, 6.8815]

[-14.5959, 22.6952]

 Robust Point Estimate

0.6110

-0.7276

0.2503

-0.4732

-1.842

4.1957

 Robust 95% CI

[-0.2223, 1.4444]

[-2.4217, 0.9665]

[-0.1492, 0.6498]

[-1.6146, 0.6681]

[-10.7285, 7.0446]

[-16.0297, 24.4211]

 Effective obs L|R

394|345

303|310

339|314

302|308

304|269

315|322

Obs L|R

1709|1610

1292|1192

1626|1584

1242|1163

1636|1583

1256|1168