out_bassefeller: DGP of Basse and Feller (2018)

In dpuelz/CliqueRT: Randomization-Based Testing of Causal Effects under General Interference

DGP of Basse and Feller (2018)

Description

Simulates an outcome vector based on the treatment assignment vector. The specific data generating procedure is adopted from Basse & Feller (2018).

Usage

out_bassefeller(
  N,
  K,
  Zobs,
  tau_main,
  housestruct = NULL,
  tau_equal = F,
  sig_c = 0.1,
  sig_y = 0.5,
  taus = 0.7,
  taup = 1.5,
  mu00 = 2,
  equal = T
)

Arguments

N	The number of observations (kids).
K	The number of houses.
Zobs	Observed treatment assignment vector.
tau_main	The main effect such that we force Y_i(1) = Y_i(0) + tau_main.
housestruct	The household structure. A vector whose length is the number of household, and where each entry specifies number of units in a household. If not specified, will generate from N, K.
tau_equal	If TRUE, then regardless of structual parameters, we set Y_i(1) = Y_i(0) + tau_main.
sig_c	Standard error on the causal effect. Here σ_μ=σ_τ=sig_c.
sig_y	Standard error on the outcome vector.
taus	Spillover effect.
taup	Primary effect.
mu00	The effect on pure control units.
equal	A binary parameter that determines the house structure when it's not specified. If equal=TRUE, every household has equal number of kids. If equal=FALSE, every household samples the number of kids at random.

Value

An outcome vector of length N.

dpuelz/CliqueRT documentation built on Jan. 6, 2023, 11:20 p.m.