RDpol: RDD poliynomial local regression

Description Usage Arguments Value

View source: R/funcion polinomios 150419.R

Description

no parametric regression discontinuity using polynomial local regression estimator with and without sample weights

Usage

1
2
3
RDpol(D = muestra1$D, Z = muestra1$x, Y = muestra1$y, c = 0,
  type = "sharp", TreatLeft = T, weights = NULL,
  svydesign_ = muestra, vcov_, kernel_ = "uniforme", h = 2, q = 2)

Arguments

D

variable with: 1 for treatment group and 0 for control group

Z

continuous variable

Y

outcome variable

c

cutpoint

type

sharp or fuzzy

TreatLeft

T if the treatment is applied for below of the cutpoint

svydesign_

design sample using survey library

vcov_

variances and covariances type: HCO, HC1, HC2 y HC3

kernel_

kernel type: triangular, epanechnikov, tricube, uniform, gaussian y cosine

h

bandwidth

q

order polynomial

W

sample weights 1/pik, NULL if there aren't sample weights

Value

Impact estimation, e.e. and t test


NataliArteaga/rddsw documentation built on Sept. 9, 2020, 4:41 p.m.