Summarize a DIF/bias screening simulation

Summarize a DIF/bias screening simulation

Usage

# S3 method for class 'mfrm_signal_detection'
summary(object, digits = 3, ...)

Arguments

object

Output from evaluate_mfrm_signal_detection().

digits

Number of digits used in numeric summaries.

...

Reserved for generic compatibility.

Value

An object of class summary.mfrm_signal_detection with:

• overview: run-level overview

• detection_summary: aggregated detection rates by design

• ademp: simulation-study metadata carried forward from the original object

• notes: short interpretation notes, including the bias-side screening caveat

See also

evaluate_mfrm_signal_detection(), plot.mfrm_signal_detection

Examples

sig_eval <- suppressWarnings(evaluate_mfrm_signal_detection(
  n_person = 20,
  n_rater = 3,
  n_criterion = 3,
  raters_per_person = 2,
  reps = 1,
  maxit = 10,
  bias_max_iter = 1,
  seed = 123
))
summary(sig_eval)
#> $overview
#> # A tibble: 1 × 5
#>   Designs Replications SuccessfulRuns ConvergedRuns MeanElapsedSec
#>     <dbl>        <dbl>          <dbl>         <dbl>          <dbl>
#> 1       1            1              1             0          0.846
#> 
#> $detection_summary
#> # A tibble: 1 × 35
#>   design_id n_person n_rater n_criterion raters_per_person DIFTargetLevel
#>   <chr>        <dbl>   <dbl>       <dbl>             <dbl> <chr>         
#> 1 S01             20       3           3                 2 C03           
#> # ℹ 29 more variables: BiasTargetRater <chr>, BiasTargetCriterion <chr>,
#> #   Reps <dbl>, ConvergenceRate <dbl>, McseConvergenceRate <dbl>,
#> #   DIFPower <dbl>, McseDIFPower <dbl>, DIFClassificationPower <dbl>,
#> #   McseDIFClassificationPower <dbl>, MeanTargetContrast <dbl>,
#> #   McseTargetContrast <dbl>, MeanTargetContrastAbs <dbl>,
#> #   McseTargetContrastAbs <dbl>, DIFFalsePositiveRate <dbl>,
#> #   McseDIFFalsePositiveRate <dbl>, BiasScreenRate <dbl>, …
#> 
#> $notes
#> [1] "Some design conditions did not converge in every replication."                                                                                                        
#> [2] "Some design conditions showed DIF power below 0.80."                                                                                                                  
#> [3] "Some design conditions showed bias-screen hit rates below 0.80."                                                                                                      
#> [4] "Some design conditions did not yield usable bias-screening t/p metrics in every replication."                                                                         
#> [5] "Bias-side rates are screening summaries derived from `estimate_bias()` output and should not be interpreted as formal power or alpha-calibrated false-positive rates."
#> [6] "MCSE columns summarize finite-replication uncertainty around the reported means and rates."                                                                           
#> 
#> $ademp
#> $ademp$aims
#> [1] "Assess DIF detection and interaction-bias screening behavior under repeated parametric many-facet simulations with known injected targets."
#> 
#> $ademp$data_generating_mechanism
#> $ademp$data_generating_mechanism$source
#> [1] "scalar_arguments"
#> 
#> $ademp$data_generating_mechanism$model
#> [1] "RSM"
#> 
#> $ademp$data_generating_mechanism$step_facet
#> [1] NA
#> 
#> $ademp$data_generating_mechanism$assignment
#> [1] "design_dependent"
#> 
#> $ademp$data_generating_mechanism$latent_distribution
#> [1] "normal"
#> 
#> $ademp$data_generating_mechanism$threshold_mode
#> [1] "generated_common"
#> 
#> $ademp$data_generating_mechanism$threshold_step_facet
#> [1] NA
#> 
#> $ademp$data_generating_mechanism$design_variables
#> [1] "n_person"          "n_rater"           "n_criterion"      
#> [4] "raters_per_person"
#> 
#> 
#> $ademp$estimands
#> [1] "DIF target-flag rate and non-target flag rate"            
#> [2] "Bias screening hit rate and screening false-positive rate"
#> [3] "Target contrast and target bias summaries"                
#> [4] "Convergence rate and elapsed time"                        
#> 
#> $ademp$methods
#> $ademp$methods$fit_method
#> [1] "JML"
#> 
#> $ademp$methods$fitted_model
#> [1] "RSM"
#> 
#> $ademp$methods$dif_method
#> [1] "residual"
#> 
#> $ademp$methods$bias_method
#> [1] "estimate_bias_screening"
#> 
#> $ademp$methods$maxit
#> [1] 10
#> 
#> $ademp$methods$quad_points
#> [1] NA
#> 
#> $ademp$methods$residual_pca
#> [1] "none"
#> 
#> 
#> $ademp$performance_measures
#> [1] "Mean detection/screening summaries across replications"
#> [2] "MCSE for means and rates"                              
#> [3] "Convergence rate"                                      
#> [4] "Bias-screen metric availability rate"                  
#> 
#> 
#> attr(,"class")
#> [1] "summary.mfrm_signal_detection"