Skip to contents

subscore

R-CMD-check

Computing Subscores in Classical Test Theory and Item Response Theory

The subscore R package provides functions for computing test subscores using several methods within the frameworks of classical test theory (CTT) and item response theory (IRT).

The package implements three major approaches to subscore computation:

  • Wainer’s augmentation method
  • Haberman’s subscore methods
  • Yen’s objective performance index (OPI)

The package also includes methods for evaluating the added value of subscores and assessing local item dependence in IRT.

Installation

You can install the development version of subscore from GitHub:

install.packages("pak")
pak::pak("daiscode/subscore")

Then load the package:

library(subscore)

Methods

Wainer’s augmentation method

The package implements Wainer’s augmentation approach for estimating subscores by incorporating information from the total test score.

Reference:

Wainer, H., Sheehan, K., & Wang, X. (2001). Some considerations when scoring subscores. In W. J. van der Linden & C. A. W. Glas (Eds.), Computerized Adaptive Testing: Theory and Practice. Kluwer Academic Publishers.

Haberman’s subscore methods

The package implements Haberman’s methods for evaluating and estimating subscores, including the computation of proportional reduction in mean squared error (PRMSE).

PRMSE provides a framework for examining whether reporting a subscore provides additional information beyond the total test score.

Reference:

Haberman, S. J. (2008). When can subscores have value? Journal of Educational and Behavioral Statistics, 33(3), 299–314. https://doi.org/10.3102/1076998607302636

Yen’s objective performance index

The package implements Yen’s objective performance index (OPI) as an approach to subscore estimation.

Reference:

Yen, W. M. (1987). A comparison of two methods for estimating subscores. ETS Research Report.

Yen’s Q3 statistic

The package also includes a function for computing Yen’s Q3 statistic, which can be used to assess the local independence assumption in item response theory.

References:

Yen, W. M. (1984). Effects of local item dependence on the fit and equating performance of the three-parameter logistic model. Applied Psychological Measurement, 8(2), 125–145. https://doi.org/10.1177/014662168400800201

Yen, W. M. (1993). Scaling performance assessments: Strategies for managing local item dependence. Journal of Educational Measurement, 30(3), 187–213. https://doi.org/10.1111/j.1745-3984.1993.tb00423.x

Main functionality

The package provides functions for:

  • Computing subscores using CTT and IRT approaches
  • Applying Wainer’s augmentation method
  • Computing Haberman’s subscore estimates
  • Computing proportional reduction in mean squared error (PRMSE)
  • Computing Yen’s objective performance index
  • Evaluating local item dependence using Yen’s Q3 statistic

Examples

See the individual function documentation for examples and usage:

?CTTsub
?subscore.s
?subscore.x
?subscore.sx
?subscore.Wainer
?subscore.corr
?Yen.OPI
?Yen.Q3

You can also see the functions included in the package with:

library(subscore)
ls("package:subscore")

Version

Current version: 3.3

Authors

Shenghai Dai — author and maintainer

Xiaolin Wang — author

Dubravka Svetina — author

License

This package is distributed under the GNU General Public License (GPL), version 2 or later.

Citation

If you use subscore in your research, please cite the package and the methodological work associated with the methods you use.