Compute the Pooled Interaction Shift Parameter Estimate From the Fold Specific Results

calc_pooled_intxn_shifts(
  intxn_shift_results,
  estimator = c("tmle", "onestep"),
  fluc_mod_out = NULL,
  a_names,
  w_names,
  z_names,
  fluctuation,
  n_folds
)

Arguments

intxn_shift_results

A list of interaction results found across the parallelized CV fold procedure

estimator

The type of estimator to be fit, either "tmle" for targeted maximum likelihood estimation or "onestep" for a one-step estimator.

fluc_mod_out

An object giving values of the logistic tilting model for targeted minimum loss estimation. This type of object should be the output of the internal routines to perform this step of the TML estimation procedure, as given by fit_fluctuation.

a_names

List of exposure names

w_names

List of covariate names

z_names

List of mediator names

fluctuation

Type of fluctuation to be used

Value

A list containing the parameter estimate, estimated variance based on the efficient influence function (EIF), the estimate of the EIF incorporating inverse probability of censoring weights, and the estimate of the EIF without the application of such weights.

Details

Estimate the value of the pooled causal parameter alongside statistical inference for the parameter estimate based on the nuisance parameters from the fold specific results for the interaction parameter.