util_varcomp_robust {dataquieR}R Documentation

Utility function to compute the rank intraclass correlation

Description

This implementation uses the package rankICC to compute the rank intraclass correlation, a nonparametric version of the ICC (Tu et al., 2023). In contrast to model-based ICC approaches, it is less sensitive to outliers and skewed distributions. It can be applied to variables with an ordinal, interval or ratio scale. However, it is not possible to adjust for covariables with this approach. The calculated ICC can become negative, like Fisher's ICC.

Usage

util_varcomp_robust(
  resp_vars = NULL,
  group_vars = NULL,
  study_data = study_data,
  meta_data = meta_data,
  min_obs_in_subgroup = 10,
  min_subgroups = 5,
  label_col = NULL
)

Arguments

resp_vars

the name of the response variable

group_vars

the name of the grouping variable

study_data

the data frame that contains the measurements

meta_data

the data frame that contains metadata attributes of study data

min_obs_in_subgroup

the minimum number of observations that is required to include a subgroup (level) of the grouping variable (group_vars) in the analysis. Subgroups with fewer observations are excluded.

min_subgroups

the minimum number of subgroups (levels) of the grouping variable (group_vars). If the variable has fewer subgroups, the analysis is not performed.

label_col

the name of the column in the metadata with labels of variables

Value

a vector from rankICC::rankICC


[Package dataquieR version 2.5.1 Index]