Summarize IDM Cut Score Objects — summary.cutsIDM • eatPrep

Builds and prints a compact summary for objects returned by computeCutsIDM.

Usage

# S3 method for class 'cutsIDM'
summary(object, digits = NULL, ...)
# S3 method for class 'summary.cutsIDM'
print(x, digits = NULL, ...)

Arguments

object: A cutsIDM object returned by computeCutsIDM.
digits: Optional integer. Number of digits used when printing numeric summary columns. The stored summary tables are not rounded.
...: Further arguments passed to print() for the individual tables.
x: A summary.cutsIDM object returned by summary().

Details

The summary method does not recompute cut scores. It reorganizes the main results already stored in the cutsIDM object:

settings reports the input format, missing-value handling, estimate and item identifier columns, number of raters, number of ordered item positions, and number of requested cuts.
boundaries connects each cut label with the numeric boundary on the internal ordinal scale and, where possible, with the lower and upper rating levels it separates.
cuts_summary contains the mean cut score on the item difficulty scale.
cut_statistics contains mean, sample standard deviation, and standard error for interpolated item positions and difficulty-scale cuts.
level_statistics describes the item-difficulty intervals implied by the mean cuts.
modal_values contains item-wise modal raw rating stages and labels.
rater_modal_correlations reports correlations between each rater's raw rating series and the item-wise modal values, with all-rater and leave-one-rater-out modal criteria.
kappa_summary and rater_kappa_statistics summarize finite pairwise Cohen's kappa values overall and per rater.
fleiss_kappa and icc_statistics provide multi-rater agreement diagnostics based on complete item rows.

The returned object is a regular list with class summary.cutsIDM, so all component tables can still be extracted programmatically.

Value

A list of class summary.cutsIDM with the components settings, boundaries, cuts_summary, cut_statistics, level_statistics, modal_values, rater_modal_correlations, kappa_summary, rater_kappa_statistics, fleiss_kappa, and icc_statistics. Printing the object returns it invisibly.

Examples

dat <- data.frame(
  est = seq(100, 500, by = 100),
  Rater1 = c(1, 1, 2, 3, 4),
  Rater2 = c(1, 2, 2, 3, 4)
)

cuts <- computeCutsIDM(dat, boundaries = c(1.5, 2.5))
summary(cuts)
#> IDM cut-score summary
#> 
#> Settings
#>  input_format missing est_col item_id_col n_raters n_items n_cuts
#>          wide    drop     est        <NA>        2       5      2
#> 
#> Boundaries
#>    cut boundary lower_level upper_level
#>  cut12      1.5           1           2
#>  cut23      2.5           2           3
#> 
#> Mean cuts on difficulty scale
#>  cut12 cut23
#>  187.5 337.5
#> 
#> Cut statistics
#>  statistic page_cut12 page_cut23 diff_cut12 diff_cut23
#>       Mean       1.88       3.37     187.50     337.50
#>         SD       0.53       0.18      53.03      17.68
#>         SE       0.38       0.13      37.50      12.50
#> 
#> Level statistics
#>  level      interval n_items mean_itemdiff sd_itemdiff
#>      1   [100,187.5)       1           100          NA
#>      2 [187.5,337.5)       2           250       70.71
#>      3   [337.5,500]       2           450       70.71
#> 
#> Modal values per item
#>  item_position item_id est n_ratings modal_n modal_prop modal_stage modal_label
#>              1       1 100         2       2        1.0           1           1
#>              2       2 200         2       1        0.5          NA        <NA>
#>              3       3 300         2       2        1.0           2           2
#>              4       4 400         2       2        1.0           3           3
#>              5       5 500         2       2        1.0           4           4
#>  modal_stages modal_labels   tie
#>             1            1 FALSE
#>           1/2          1/2  TRUE
#>             2            2 FALSE
#>             3            3 FALSE
#>             4            4 FALSE
#> 
#> Rater correlations with modal values
#>  person n_items_modal_all cor_modal_all n_items_modal_loo
#>  Rater1                 4             1                 5
#>  Rater2                 4             1                 5
#>  cor_modal_leave_one_out
#>                     0.94
#>                     0.94
#> 
#> Pairwise Cohen kappa summary
#>  n_pairs mean_kappa sd_kappa mean_n_items
#>        1       0.74       NA            5
#> 
#> Rater pairwise Cohen kappa
#>  person n_pairs mean_kappa sd_kappa mean_n_items
#>  Rater1       1       0.74       NA            5
#>  Rater2       1       0.74       NA            5
#> 
#> Fleiss kappa
#>  method n_items n_raters kappa statistic p_value
#>  Fleiss       5        2  0.73      2.79    0.01
#> 
#> ICC agreement and consistency
#>         type  model   unit n_items n_raters icc_name  icc f_value df1 df2
#>    agreement twoway single       5        2 ICC(A,1) 0.93      29   4   4
#>  consistency twoway single       5        2 ICC(C,1) 0.93      29   4   4
#>  p_value conf_level lbound ubound
#>        0       0.95   0.59   0.99
#>        0       0.95   0.50   0.99

sum_cuts <- summary(cuts)
sum_cuts$boundaries
#> # A tibble: 2 × 4
#>   cut   boundary lower_level upper_level
#>   <chr>    <dbl> <chr>       <chr>      
#> 1 cut12      1.5 1           2          
#> 2 cut23      2.5 2           3          
sum_cuts$level_statistics
#> # A tibble: 3 × 5
#>   level interval      n_items mean_itemdiff sd_itemdiff
#>   <int> <chr>           <int>         <dbl>       <dbl>
#> 1     1 [100,187.5)         1           100        NA  
#> 2     2 [187.5,337.5)       2           250        70.7
#> 3     3 [337.5,500]         2           450        70.7
sum_cuts$rater_kappa_statistics
#> # A tibble: 2 × 5
#>   person n_pairs mean_kappa sd_kappa mean_n_items
#>   <chr>    <int>      <dbl>    <dbl>        <dbl>
#> 1 Rater1       1      0.737       NA            5
#> 2 Rater2       1      0.737       NA            5
sum_cuts$icc_statistics
#> # A tibble: 2 × 14
#>   type   model unit  n_items n_raters icc_name   icc f_value   df1   df2 p_value
#>   <chr>  <chr> <chr>   <int>    <int> <chr>    <dbl>   <dbl> <dbl> <dbl>   <dbl>
#> 1 agree… twow… sing…       5        2 ICC(A,1) 0.933    29.0     4     4 0.00326
#> 2 consi… twow… sing…       5        2 ICC(C,1) 0.933    29.0     4     4 0.00326
#> # ℹ 3 more variables: conf_level <dbl>, lbound <dbl>, ubound <dbl>