Mar 27, 2019Mar 29, 2019
8:30 am4:30 pm
Thurgood Marshall Room, 2113 Stamp Student Union
HDQM

The Center for Integrated Latent Variable Research (CILVR) presents Introduction to Item Response Theory, taught by Ji Seung Yang.

A central theme in psychological and educational measurement is the establishment of the technical criteria and appropriate use of statistical models for ensuring the reliability, fairness, and validity of the scores that are obtained via assessments and surveys. From large-scale educational assessment (e.g., the Graduate Record Examinations; GRE) to the assessment of physical, mental, and social well-being for patients (e.g., Patient Reported Outcomes Measurement Information System), the underlying principles of measurement are similar. Item Response Theory (IRT) refers to a particular set of the modern measurement theory, models, and techniques that are utilized to develop either a large-scale assessment or a survey.  

The purpose of this three-day course is to introduce IRT to participants who have not been formally exposed to it or who would like to consolidate/extend their previous knowledge by covering both theoretical and practical basics of various IRT models including their estimation, assessment of model fit, and application to real-world assessments. The course materials were developed in collaboration with Dr. Yang Liu (also at the University of Maryland) while the short course delivery will be made by Dr. Ji Seung Yang.

The course starts with a brief review of the history of IRT. Then IRT models for various item responses (e.g., dichotomous and polytomous item responses; 1, 2, 3-parameter models, graded response models, partial credit models, and nominal response models) are introduced with empirical/simulated data examples. Estimation of model parameters and related issues are discussed. Various approaches to assess model fit are addressed before moving into more applied topics that include detection of local dependence and differential item functioning. Overall this course focuses on the statistical theory underlying IRT models as well on their applications to real and simulated data sets using computer software packages.

Empirical examples and hands-on exercises for this workshop will be done using the flexMIRT software and other supplemental software packages such as R and Excel. Participants are strongly recommended to bring the package loaded on their own laptop for a more enhanced learning experience. Further instructions will be distributed and no additional software cost will be needed for this course.

After completing the workshop, participants will be able to apply appropriate unidimensional IRT models to their own empirical data set to analyze psychometric properties of test or survey items, to obtain individual scale scores, to assess model fit and assumptions, and to detect differential item functioning items using IRT models.

It is assumed that participants have knowledge of general and generalized linear models, especially logistic regression (e.g., hypothesis testing, confidence intervals, parameter interpretation, and likelihood ratio tests). Although not required, a participant’s experience in this workshop will be enhanced by additional prior coursework or experience with measurement theory (e.g., Classical Test Theory) and latent variable modeling such as factor analysis.

Empirical examples and hands-on exercises for this workshop will be done using the flexMIRT software (to which on-site participants will have temporary access) and other supplemental software packages such as R and Excel. Participants are strongly recommended to bring the R package loaded on their own laptop for a more enhanced learning experience. On-line participants may use whatever software they have access to on their own (e.g., R, flexMIRT).

Schedule
Check-in: 8:30 AM
Continental Breakfast*: 8:30 AM - 9:00 AM
Morning Session: 9:00 AM - 12:15 PM
Lunch (on your own): 12:15 PM - 1:15 PM
Afternoon Session: 1:15 PM - 4:30 PM
* Participants who have special dietary needs or preferences are welcome to bring their own food as well.

Course Fees: $895 for all three days; $595 for full-time students; (free for registered HDQM Department faculty and students, although you must register; admission will depend on if space is available); $295 for all three days with the online option.

Register today using the Short-Course Registration button! For students not in the HDQM department, please fill out the Item Response Theory Workshop Registration Form and upload it when prompted within the online registration system:
Item Response Theory Workshop Registration Form