ATEcutoff: Create ATE Cutoff

In zieglerjef/openEnded: Tools for Implementing and Analyzing Open-Ended Responses as Manipulation Checks

Description

Create cutoff, then bootstrap or simulate the sampling distribution of participants that likely received the treatment (ATE among "compliers" and "non-compliers").

Usage

ATEcutoff(
  dataframe,
  formula,
  similarity_measures,
  type_model,
  k,
  bounds,
  user_seed,
  plot_treatment = NULL,
  plot_interact_x = NULL,
  stable_x
)

Arguments

dataframe	Dataframe from which we will estimate our regression model.
formula	Symbolic representation of the model to be estimated. This is written in "typical" R language (i.e. y ~ x1 + x2), such that y is the outcome variable and x1 and x2 are the predictors.
similarity_measures	Vector(s) from dataframe that contains the similarity measures to be used as weights. Possible values for measure_type = c("osa", "lv", "dl", "hamming", "lcs", "qgram", "cosine", "jaccard", "jw", "soundex"). Default is "jaccard".
type_model	Statistical model to estimate. Currently support OLS and logistic ("ls", "logit").
k	The penalty that you want to set for down-weighting inattentive respondents. Lower levels of k down-weight low attention participants more severely.
bounds	Minimum and maximum of uniform distribution we should draw cutoff values between.

Value

Dataframe of all estimated marginal effects that make up the sampling distribution for participants that did and did not receive the treatment.

Author(s)

Jeffrey Ziegler (<jeffrey.ziegler[at]emory.edu>)

See Also

plotComplierATE

Examples

ATEcutoff(dataframe=replication_complete.cases, similarity_measures=c("jaccardDist", "cosineDist"), bounds=0.2, formula=trustChurch_postTreat ~ Concordant*attendanceBin, model_type="ls", k=3))

zieglerjef/openEnded documentation built on Nov. 30, 2020, 2:03 p.m.