Software & Code

R Packages

personnelSelectionUtility

personnelSelectionUtility implements classical and contemporary utility-analysis methods for personnel selection under a unified notation, organized by criterion scale (classificatory or continuous/monetary) and selection structure (compensatory or multiple-hurdle). The package covers Taylor–Russell univariate and Thomas–Owen–Gunst multivariate classification models, Brogden–Cronbach–Gleser and Boudreau-style continuous and monetary utility, Schmidt–Hunter–Pearlman intervention utility, Sturman-style incremental validity for multiple predictors and outcomes, simulation tools for compensatory and staged multiple-hurdle systems, Pareto frontiers for validity–diversity trade-offs, AUC-to-effect-size conversions, and Monte Carlo uncertainty propagation.

Package website GitHub

Installation:

# From CRAN
install.packages("personnelSelectionUtility")

# Development version
remotes::install_github("rgempp/personnelSelectionUtility", build_vignettes = TRUE)

csemGT

csemGT estimates conditional standard errors of measurement (CSEMs) under Generalizability Theory for the univariate single-facet design in which persons are crossed with items (a p × i design). Whereas a single overall SEM summarises measurement precision for a whole test, a conditional SEM describes how precision varies along the score scale, so that examinees at different observed-score levels can be assigned different error bands.

The package implements the absolute-error estimator and the three relative-error estimators (full, large_a, uncorrelated) developed by Brennan (1998), together with optional polynomial smoothing of the conditional error variances across the score scale and analytical or bootstrap standard errors for the per-person estimates. csemGT is the R companion to the Stata gtcsem command and reproduces its numerical results.

Package website GitHub

Installation:

# From CRAN
install.packages("csemGT")

# Development version
remotes::install_github("rgempp/csemGT", build_vignettes = TRUE)

Stata Packages

gtcsem

gtcsem is a Stata module that estimates per-person conditional standard errors of measurement (CSEMs) under Generalizability Theory for the univariate, single-facet, persons-by-items (p × i) crossed design. Unlike the overall (population-level) SEM, which summarizes measurement precision averaged across persons, the CSEM characterizes the precision of measurement for an individual at a given score level, the relevant index when decisions concern individuals rather than groups, including the local accuracy of cut-score classifications. The package implements three estimators of the per-person relative error variance (Brennan, 1998), analytical and bootstrap standard errors, generalizability and dependability coefficients, D-study extrapolation to hypothetical test lengths, and a companion command (gtcsem_plot) for Brennan-style CSEM plots.

GitHub

Installation:

# From SSC
ssc install gtcsem

# From GitHub (development version)
net install gtcsem, from("https://raw.githubusercontent.com/rgempp/gtcsem/main/") replace

Requires: Stata 16 or later.

Legacy Programs (Delphi)

More years ago than I care to remember, I developed a few small programs in Delphi for teaching purposes. Here are the ones that survived successive computer upgrades. They are in Spanish, Windows-only, and come with no documentation, but they still work.

  • PhiLambda (philambda.exe): A small utility for estimating the PhiLambda coefficient (reliability of a dichotomous decision in Generalizability Theory), based on the ANOVA table from the RELIABILITIES menu in SPSS.
  • WinError (WinError.exe): An interactive tool for explaining what the standard error of measurement is and what it is used for.
  • Notas (notas.exe): A quick converter from percentage scores to Chilean grades (1 to 7 scale).