R: Population-Weighted Heterogeneous Effects Analysis of...

pIMCE {cjbart}

R Documentation

Population-Weighted Heterogeneous Effects Analysis of Conjoint Results

Description

pIMCE calculates the population individual-level marginal component effects from a BART-estimated conjoint model, using marginal attribute distributions specified by the researcher.

Usage

pIMCE(
  model,
  covar_data,
  attribs,
  l,
  l_1,
  l_0,
  marginals,
  method = "bayes",
  alpha = 0.05,
  cores = 1,
  skip_checks = FALSE,
  verbose = TRUE
)

Arguments

`model`	A model object, the result of running `cjbart()`
`covar_data`	A data.frame of covariate information to predict pIMCEs over
`attribs`	Vector of attribute names
`l`	Name of the attribute of interest
`l_1`	Attribute-level of interest for attribute l
`l_0`	Reference level for attribute l
`marginals`	A named list where every element is a named vector of marginal probabilities for each corresponding attribute-level. For example, `marginals = list("A1" = c("q" = 0.4, "r" = 0.6), "A2" = c("x" = 0.7, "y" = 0.2, "z" = 0.1))`
`method`	Character string, setting the variance estimation method to use. When method is "parametric", a typical combined variance estimate is employed; when `method = "bayes"`, the 95% posterior interval is calculated; and when `method = "rubin"`, combination rules are used to combine the variance analogous to in multiple imputation analysis.
`alpha`	Number between 0 and 1 – the significance level used to compute confidence/posterior intervals. When `method = "bayes"`, the posterior interval is calculated by taking the alpha/2 and (1-alpha/2) quantiles of the posterior draws. When `method = "rubin"`, the confidence interval equals the IMCE +/- `qnorm(alpha/2)`. By default, alpha is 0.05 i.e. generating a 95% confidence/posterior interval.
`cores`	Number of CPU cores used during prediction phase
`skip_checks`	Boolean, indicating whether to check the structure of the data (default = `FALSE`). Only set this to `TRUE` if you are confident that the data is structured appropriately
`verbose`	Boolean, indicating whether to print progress (default = TRUE)

Details

This function calculates the population-weighted IMCE, which takes into account the population distribution of profiles. Rather than average over the multiple OMCE estimates, this function generates estimated treatment effects for all possible potential outcomes along all attributes except the attribute of interest, and then marginalizes these over the supplied marginal distributions. Uncertainty estimates are recovered using credible intervals.

Value