This function implements the bias correction of the ranking distribution using a paired anchor question, using the IPW estimator.
Usage
imprr_weights(
data,
J = NULL,
main_q,
anc_correct,
population = "non-random",
assumption = "contaminated",
seed = 123456,
weight = NULL,
ranking = "ranking"
)Arguments
- data
The input dataset with ranking data.
- J
The number of items in the ranking question. Defaults to NULL, in which case it will be inferred from the data, only if the column for `main_q` exists in the data.
- main_q
Column name for the main ranking question to be analyzed. Using this argument, the function automatically looks for columns with marginal rankings. For example, if `main_q` is `app_identity`, the function looks for `app_identity_1`, `app_identity_2`, `app_identity_3`, and so on, with an underbar separator followed by numbers.
- anc_correct
Indicator for passing the anchor question.
- population
Choice of the target population out of non-random respondents (default) or all respondents.
- assumption
Choice of the identifying assumption if `population` is set to all
- seed
Seed for
set.seedfor reproducibility.- weight
A vector of weights. Defaults to NULL.
- ranking
The name of the column that will store the full ranking profile. Defaults to "ranking". If `main_q` exists in the data, the produced column should be identical to `main_q`. However, the function defaults to creating another column by combining marginal rankings, just in case.