Back to Tristan Zajonc's page.

OpenIRT - Bayesian and Maximum Likelihood Estimation of Item Response Theory (IRT) models in Stata.

OpenIRT estimates 2PL and 3PL Item Response Theory (IRT) models for dichotomous data in Stata. It includes Bayesian MCMC estimation of item parameters and abilities, and maximum likelihood ability estimates. The Bayesian methods includes both expected posterior (EAP) and plausible values (PV) ability and item parameter estimates.

OpenIRT is designed to be flexible. It allows any combination of items, abilities, and responses to be missing. For instance, fixed anchor items can be used to place children and items on a known scale (e.g. TIMSS or NAEP); missing responses can be used to link multiple overlapping test forms; and known abilities can be used to calibrate new items and ability parameters.

OpenIRT relies on compiled code that works on Windows, Mac, and Linux. It requires at least Stata 10.1.

Last updated: January 10, 2011.

Install in Stata

net install http://www.people.fas.harvard.edu/~tzajonc/stata/openirt/openirt

Help

help openirt

Example: Estimate both item parameters and ability for single test

openirt, id(id) item_prefix(item) save_item_parameters("items.dta") save_trait_parameters("traits.dta")

Slides, Sample Code, Source

From World Bank Workshop (December 9, 2009):

• Slides on IRT and OpenIRT [pdf]
• Example code: openirt_examples.do. Includes examples of plotting item characteristic curves, item information curves, and test information and characteristic curves.
• Source: OpenIRT on BitBucket

Motivating paper

India shining and Bharat drowning: comparing two Indian states to the worldwide distribution in mathematics achievement (with Jishnu Das) Journal of Development Economics. forthcoming.

Problems?

If you experience problems, please email me: tzajonc@fas.harvard.edu.