itemrest: Evaluate Item Removal Strategies for Exploratory Factor...

In ItemRest: Automated Item Removal Strategies for Exploratory Factor Analysis

Evaluate Item Removal Strategies for Exploratory Factor Analysis (EFA)

Description

This function automates the process of identifying low-quality items (those with low factor loadings or significant cross-loadings) in an Exploratory Factor Analysis (EFA). It systematically tests various combinations of removing these problematic items and evaluates the impact on model fit, returning a comprehensive summary of all tested strategies.

Usage

itemrest(
  data,
  cor_method = "pearson",
  n_factors = NULL,
  extract = "uls",
  rotate = "oblimin",
  min_loading = 0.3,
  loading_diff = 0.1
)

Arguments

data	A numeric data.frame or matrix for the analysis.
cor_method	The correlation method to use, e.g., "pearson" or "polychoric".
n_factors	The number of factors. If NULL, it's determined automatically by parallel analysis.
extract	The factor extraction (estimation) method. See psych::fa. Default is "uls".
rotate	The rotation method. See psych::fa. Default is "oblimin".
min_loading	Minimum factor loading threshold. Default is 0.30.
loading_diff	Threshold for cross-loading identification. Default is 0.10. Can be "howard".

Value

An object of class itemrest_result. This is a list containing the following components:

descriptive_stats	Basic descriptive statistics of the input data.
initial_efa	The results of the initial EFA before any items are removed.
problem_items	A list of items identified as low-loading or cross-loading.
removal_summary	A data.frame summarizing the results of all tested removal strategies.
optimal_strategy	The best-performing strategy that resulted in a clean factor structure (no cross-loadings).
settings	A list of the settings used for the analysis.

Examples

# We will use the 'bfi' dataset from the 'psych' package.
# This requires the 'psych' package to be installed.
if (requireNamespace("psych", quietly = TRUE)) {
  data(bfi, package = "psych")

  # 1. Prepare the data: Select the personality items (first 25 columns)
  #    and remove rows with missing values for this example.
  example_data <- bfi[, 1:25]
  example_data <- na.omit(example_data)

  # 2. Run the item removal analysis.
  #    Based on theory, the Big Five model has 5 factors.
  results <- itemrest(
    data = example_data,
    n_factors = 5,
    cor_method = "pearson" # Data is not ordinal, so pearson is appropriate
  )

  # 3. Print the report for optimal strategies (default).
  print(results)

  # 4. Print the report for all tested strategies.
  print(results, report = "all")
}

ItemRest documentation built on April 13, 2026, 5:07 p.m.