Sage Journals: Discover world-class research

Abstract

The R package autoRasch has been developed to perform a Rasch analysis in a (semi-)automated way. The automated part of the analysis is achieved by optimizing the so-called in-plus-out-of-questionnaire log-likelihood (IPOQ-LL) or IPOQ-LL-DIF when differential item functioning (DIF) is included. These criteria measure the quality of fit on a pre-collected survey, depending on which items are included in the final instrument. To compute these criteria, autoRasch fits the generalized partial credit model (GPCM) or the generalized partial credit model with differential item functioning (GPCM-DIF) using penalized joint maximum likelihood estimation (PJMLE). The package further allows the user to reevaluate the output of the automated method and use it as a basis for performing a manual Rasch analysis and provides standard statistics of Rasch analyses (e.g., outfit, infit, person separation reliability, and residual correlation) to support the model reevaluation.

Keywords

Rasch analysis semi-automated analysis PJMLE coordinate descent lasso penalty

Description of the Package

Rasch analysis is a popular statistical method for developing and validating instruments to measure psychological aspects of humans. Typically, this method is done manually by experts with the assistance of software packages such as WINSTEP, RUMM2030, ConQuest, or eRM (www.rasch.org, 2014). Initially, the responses from an original survey need to be fitted to the model, either with the Rasch model (Rasch, 1960) in dichotomous cases or with the partial credit model (PCM) (Masters, 1982) in polytomous cases. As Rasch properties are checked (e.g., item goodness-of-fit, local dependency, reliability, and unidimensionality), items that misfit are removed one by one manually in an iterative process (Mul et al., 2021; O’Brien et al., 2021). Therefore, this procedure can be relatively time-consuming, especially when dealing with large and complex datasets. In addition, since the decisions on which items to include are partly based on human judgments combined with clinical expertise, different experts often select instruments that are different but equally suited.

The R (R Core Team, 2021) package autoRasch is a new tool for performing Rasch analysis in a (semi-)automatic manner. This package implements novel criteria that have been shown to naturally incorporate the standard properties of Rasch analysis (e.g., item goodness-of-fit, local dependency, reliability, unidimensionality, and differential item functioning). These criteria are called the in-plus-out-of-questionnaire log-likelihood with differential item functioning (IPOQ-LL-DIF) for cases incorporating DIF and in-plus-out-of-questionnaire log-likelihood (IPOQ-LL) for non-DIF cases (Wijayanto et al., 2021, 2022).

The IPOQ-LL-DIF and IPOQ-LL are computed by implementing the generalized partial credit model with differential item functioning (GPCM-DIF) (Wijayanto et al., 2022) and the generalized partial credit model (GPCM) (Muraki, 1992), respectively. For simplicity, the autoRasch package implements penalized joint maximum likelihood estimation (PJMLE) to fit these models. Moreover, recent studies have shown that PJMLE could yield comparable estimates to the marginal maximum likelihood estimation (MMLE) (Chen et al., 2019; Paolino, 2013; Robitzsch, 2021).

In contrast to the standard Rasch analysis, which investigates and removes items in a manual iterative way using statistics and expert knowledge, the autoRasch package attempts to automate the process by maximizing either the IPOQ-LL-DIF or IPOQ-LL scoring criteria. To maximize the score, this package implements a stepwise selection search method. The output of the stepwise selection search is formatted in a way that makes it easy to read, and it contains relevant information (e.g., the highest score and the items that yield it). Additionally, the change in score when sequentially removing items can be observed visually.

As a semi-automated method, our new procedure always welcomes the application of expert knowledge, for example, through pre- and post-analysis. To facilitate these pre- and post-analyses, the autoRasch package implements the fitting functions for the partial credit model (PCM) (Masters, 1982) and the partial credit model with differential functioning (PCM-DIF) (Wijayanto et al., 2022). Some standard Rasch statistics (e.g., outfit and infit statistics (Wright & Masters, 1982), residual correlation (Marais, 2013), and person/item separation reliability (PSR/ISR) (Wright & Masters, 1982) are also implemented. In addition, some typical graphical tools are also available (e.g., person-item map, item characteristic curve (ICC), and expected value curve (EVC)).

Availability, Documentation, and Distribution

The autoRasch package, user’s manual, and sample codes are available from Comprehensive R Archive Network (CRAN) repository at https://CRAN.R-project.org/package=autoRasch. The development version of the package is hosted on GitHub (https://github.com/fwijayanto/autoRasch).

Supplemental Material

Supplemental Material - autoRasch: An R Package to Do Semi-Automated Rasch Analysis

Supplemental Material for autoRasch: An R Package to Do Semi-Automated Rasch Analysis by Feri Wijayanto, Ioan Gabriel Bucur, Perry Groot, and Tom Heskes in Applied Psychological Measurement

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research leading to these results has received funding from Universitas Islam Indonesia.

ORCID iD

Feri Wijayanto

Supplemental Material

Supplemental material for this article is available online.

References

Chen

Zhang

(2019). Joint maximum likelihood estimation for high-dimensional exploratory item factor analysis. Psychometrika, 84(1), 124–146. https://doi.org/10.1007/s11336-018-9646-5

Marais

(2013). Local dependence. In Christensen

K. B.

Kreiner

Mesbah

(Eds.), Rasch models in health chapter 7 (pp. 111–128). Wiley.

Masters

G. N.

(1982). A rasch model for partial credit scoring. Psychometrika, 47(2), 149–174. https://doi.org/10.1007/bf02296272

Mul

Hamadeh

Horlings

C. G.

Tawil

Statland

J. M.

Sacconi

Corbett

A. J.

Voermans

N. C.

Faber

C. G.

van Engelen

B. G.

Merkies

I. S.

(2021). The facioscapulohumeral muscular dystrophy Rasch-built overall disability scale (FSHD-RODS). European Journal of Neurology, 28(7), 2339–2348. https://doi.org/10.1111/ene.14863

Muraki

(1992). A generalized partial credit model: Application of an EM algorithm. Applied Psychological Measurement, 16(2), 159–176. https://doi.org/10.1177/014662169201600206

O’Brien

K. K.

Dzingina

Harding

Gao

Namisango

Avery

Davis

A. M.

(2021). Developing a short-form version of the HIV disability questionnaire (SF-HDQ) for use in clinical practice: A Rasch analysis. Health and Quality of Life Outcomes, 19(1), 1–21. https://doi.org/10.1186/s12955-020-01643-2

Paolino

J.-P.

(2013). Penalized joint maximum likelihood estimation applied to two parameter logistic item response models [PhD thesis, Columbia University].

Rasch

(1960). Studies in mathematical psychology: I. Probabilistic models for some intelligence and attainment tests. Nielsen & Lydiche.

R Core Team (2021). R: A language and environment for statistical computing. R Foundation for Statistical Computing.

10.

Robitzsch

(2021). A comprehensive simulation study of estimation methods for the Rasch model. Stats, 4(4), 814–836. https://doi.org/10.3390/stats4040048

11.

Wijayanto

Bucur

I. G.

Mul

Groot

van Engelen

B. G.

Heskes

(2022). Semi-automated Rasch analysis with differential item functioning. PsyArxiv preprint.

12.

Wijayanto

Mul

Groot

van Engelen

B. G.

Heskes

(2021). Semi-automated Rasch analysis using in-plus-out-of-questionnaire log likelihood. The British Journal of Mathematical and Statistical Psychology, 74(2), 313–339. https://doi.org/10.1111/bmsp.12218

13.

Wright

B. D.

Masters

G. N.

(1982). Rating scale Analysis. MESA Press.

14.

www.rasch.org (2014). Rasch measurement analysis software directory.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.58 MB