A full Bayesian implementation of a generalised partial credit model with an application to an international disability survey
A full Bayesian implementation of a generalised partial credit model with an application to an international disability survey
Generalised partial credit models (GPCM) are ubiquitous in many applications in the health and medical sciences that use item response theory. Such polytomous item response models have a great many uses ranging from assessing and predicting an individual's latent trait to ordering the items to test the effectiveness of the test instrumentation. By implementing these models in a full Bayesian framework, computed through the use of Markov chain Monte Carlo (MCMC) methods implemented in the efficient STAN software package, this article exploits the full inferential capability of the GPCMs. The GPCMs include explanatory covariate effects which allow simultaneous estimation of regression and item parameters. The Bayesian methods for ranking the items using the Fisher information criterion (FIC) are implemented using MCMC. This allows us to fully propagate and ascertain uncertainty in the inferences by calculating the posterior predictive distribution of item specific FIC in a novel manner that has not been exploited in the literature before. Lastly, we propose a new Monte Carlo method for predicting the latent trait score of a new individual by approximating the relevant Bayesian predictive distribution. Data from a Model Disability Survey carried out in Sri Lanka by the World Health Organisation (WHO) and the World Bank are used to illustrate the methods. The proposed approaches are shown to provide simultaneous model based inference for all aspects of disability which can be explained by environmental and socio-economic factors.
Sahu, Sujit
33f1386d-6d73-4b60-a796-d626721f72bf
Bass, Mark
53ed9f53-65a0-41bf-904b-1f4e73cc8365
Sabariego, Carla
5caee823-f6fd-45a7-bbbb-990459603d27
Cieza, Alarcos
030f863e-7491-4d69-a195-52a765ae443f
Fellinghauer, Carolina
194b61ef-c33b-4d80-99c3-07d30a92b60b
Chatterji, Somnath
3cb0490f-e7c3-408b-a9da-37ccfd3e070d
Sahu, Sujit
33f1386d-6d73-4b60-a796-d626721f72bf
Bass, Mark
53ed9f53-65a0-41bf-904b-1f4e73cc8365
Sabariego, Carla
5caee823-f6fd-45a7-bbbb-990459603d27
Cieza, Alarcos
030f863e-7491-4d69-a195-52a765ae443f
Fellinghauer, Carolina
194b61ef-c33b-4d80-99c3-07d30a92b60b
Chatterji, Somnath
3cb0490f-e7c3-408b-a9da-37ccfd3e070d
Sahu, Sujit, Bass, Mark, Sabariego, Carla, Cieza, Alarcos, Fellinghauer, Carolina and Chatterji, Somnath
(2019)
A full Bayesian implementation of a generalised partial credit model with an application to an international disability survey.
Journal of the Royal Statistical Society, Series C (Applied Statistics).
(doi:10.1111/rssc.12385).
Abstract
Generalised partial credit models (GPCM) are ubiquitous in many applications in the health and medical sciences that use item response theory. Such polytomous item response models have a great many uses ranging from assessing and predicting an individual's latent trait to ordering the items to test the effectiveness of the test instrumentation. By implementing these models in a full Bayesian framework, computed through the use of Markov chain Monte Carlo (MCMC) methods implemented in the efficient STAN software package, this article exploits the full inferential capability of the GPCMs. The GPCMs include explanatory covariate effects which allow simultaneous estimation of regression and item parameters. The Bayesian methods for ranking the items using the Fisher information criterion (FIC) are implemented using MCMC. This allows us to fully propagate and ascertain uncertainty in the inferences by calculating the posterior predictive distribution of item specific FIC in a novel manner that has not been exploited in the literature before. Lastly, we propose a new Monte Carlo method for predicting the latent trait score of a new individual by approximating the relevant Bayesian predictive distribution. Data from a Model Disability Survey carried out in Sri Lanka by the World Health Organisation (WHO) and the World Bank are used to illustrate the methods. The proposed approaches are shown to provide simultaneous model based inference for all aspects of disability which can be explained by environmental and socio-economic factors.
Text
bayesian_gpcm
- Accepted Manuscript
More information
Accepted/In Press date: 16 September 2019
e-pub ahead of print date: 28 October 2019
Identifiers
Local EPrints ID: 434331
URI: http://eprints.soton.ac.uk/id/eprint/434331
ISSN: 0035-9254
PURE UUID: a96414d5-e5b0-4a8c-b051-e6e1002f48cd
Catalogue record
Date deposited: 19 Sep 2019 16:30
Last modified: 17 Mar 2024 02:51
Export record
Altmetrics
Contributors
Author:
Mark Bass
Author:
Carla Sabariego
Author:
Alarcos Cieza
Author:
Carolina Fellinghauer
Author:
Somnath Chatterji
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics