Skip to contents

Item Restscore Analysis

Source: R/item_restscore.R
RMitemrestscore.Rd

Computes observed and model-expected item-restscore correlations using iarm::item_restscore(), and enriches the output with the absolute difference between observed and expected values, item average locations, and item locations relative to the sample mean person location.

Usage

RMitemRestscore(data, output = "kable", sort, p_adj = "BH")

Arguments

data

A data.frame or matrix of item responses. Items must be scored starting at 0 (non-negative integers). Missing values (NA) are allowed, but at least one complete case (row with no NA) must be present.

output

Character string controlling the return value. Either "kable" (default) for a formatted knitr::kable() table, or "dataframe" for the underlying data.frame.

sort

Optional character string. When sort = "diff", rows are sorted by the absolute magnitude of Difference in descending order, so that both over- and underfitting items appear near the top.

p_adj

Character string specifying the p-value adjustment method passed to iarm::item_restscore(). Default "BH" (Benjamini-Hochberg); use "none" for unadjusted p-values. Run ?stats::p.adjust for the list of available methods.

Value

  • If output = "kable": a knitr_kable object (plain text table via format = "pipe") with columns for item name, observed and expected restscore correlations, the signed difference (observed minus expected), adjusted p-value, the Flagged misfit label, and item location relative to the sample mean person location.

  • If output = "dataframe": a data.frame with columns Item, Observed, Expected, Difference, p_adjusted, Flagged, and Relative_location. Flagged is "overfit" (observed above expected, adj. p < .05), "underfit" (below, adj. p < .05), or "" (not flagged).

The Difference column is signed (observed minus expected): positive values indicate that the item correlates more strongly with the rest-score than the Rasch model predicts (over-discrimination / overfit, often associated with local dependence), and negative values indicate weaker-than-expected association (under-discrimination / underfit, often associated with multidimensionality or noise).

Details

Item-restscore correlations using Goodman-Kruskal's gamma (Kreiner, 2011) measure the association between a person's score on a single item and their total score on the remaining items (the "restscore"). Under a correctly fitting Rasch model, observed and model-expected correlations should agree closely.

Item parameters are estimated by conditional maximum likelihood via psychotools::pcmodel() (a dichotomous item is a 2-category PCM); the item-restscore statistic itself comes from iarm::item_restscore() and is conditional on the total score, so it is invariant to the estimation engine. Per-item average locations are the means of the CML thresholds, and the person-location reference is the mean of the Warm WLE estimates.

Relative item location is defined as the item's average location minus the sample mean person location, providing a measure of item targeting.

The iarm package must be installed (it is in Suggests, not Imports).

References

Kreiner, S. (2011). A Note on Item–Restscore Association in Rasch Models. Applied Psychological Measurement, 35(7), 557–561. doi:10.1177/0146621611410227

Examples

# \donttest{
if (requireNamespace("iarm", quietly = TRUE)) {
  # Simulate binary item response data (8 items, 200 persons)
  set.seed(42)
  sim_data <- as.data.frame(
    matrix(sample(0:1, 200 * 8, replace = TRUE), nrow = 200, ncol = 8)
  )
  colnames(sim_data) <- paste0("Item", 1:8)

  # Default kable output
  RMitemRestscore(sim_data)

  # Sorted by absolute difference
  RMitemRestscore(sim_data, sort = "diff")

  # Return as data.frame for further processing
  df <- RMitemRestscore(sim_data, output = "dataframe")
}
#> 
#> 
#> 
# }