Introduction to classical test theory, item response models

Analysis of Measurement Instruments:
Introduction to classical test theory, item response models and multilevel item
response models
Twente University
Dates: 16 – 19 December 2014
Target
The course provides an introduction on classical and modern test theory and its major applications, such
as the construction of measurement instruments, linking and equating measurements, evaluation
differential item functioning, optimal test construction, computerized adaptive testing and the
evaluation of large-scale educational surveys such as PISA and TIMSS. The introduction is broad and on a
conceptual level, but references will be provided to the more technical aspects such as estimation and
testing of the models.
The second part of the course will focus on Bayesian item response modeling. An introduction will be
given to the important features of the Bayesian modeling approach, which will be illustrated with real
data examples using software packages OpenBugs and R. The practicals will focus on standard software,
but interested students will also be given the opportunity to get acquainted with Bayesian modeling
software.
Tutors
Prof. Dr. Cees Glas & Dr. Hanneke Geerlings
Maximum number of participants: 25
Venue
Twente University, Enschede.
Prior Requirements
Knowledge of statistics and methodology in the field of the behavioral and social sciences
Literature
Brennan, R. L. (1992). Generalizability theory. NCME Instructional Module 14. Public Domain.
C.A.W. Glas (2010). Introduction to IRT. Handout
C.A.W. Glas (2010). Manual MIRT. Public domain, handout.
Holman, R., & Glas, C. A. W. (2005). Modeling non-ignorable missing data mechanisms with item
response theory models. British Journal of Mathematical and Statistical Psychology. 58, 1-17.
Korobko, O.B., Glas, C.A.W., Bosker, R.J. & Luyten, J.W. (2008). Comparing the difficulty of examination
subjects with item response theory. Journal of Educational Measurement, 45, 139–157.
Kim, S.-H. (2001). An evaluation of a Markov Chain Monte Carlo method for the Rasch model. Applied
Psychological Measurement, 25, 163-176.
Albert, J.H. (1992). Bayesian estimation of normal ogive item response functions using Gibbs sampling.
Journal of Educational Statistics, 17, 251-269.
Work method
The course takes 4 full days. Every session consists of a lecture of approximately 2.5 hours and a
computer workshop of approximately 3 hours.
Registration
Please send an email to [email protected]
Participants will be accepted by order of application date, the number of participants is limited to 25
persons.
Accommodation
There are various hotels in Enschede, some are located near the University of Twente
Course fee and payment
The course is free for (aspirant) IOPS students. For non-IOPS students the course fee is € 150,-, an invoice
will be sent. Reductions or refunds are not possible.
Travel costs, accommodation and meals are not included. Please make your own arrangements.
Contents
In general, the course provides an overview of classical and modern test theory discusses the coherence
of the various theories and deals with the application of test theory for developing and evaluating
measurement instruments for the social sciences.
Lecture 1
Subjects
Workshop 1
Literature
Introduction
Overview of classical test theory and item response theory, the relation between the two
approaches and some important applications such as evaluation of reliability and validity,
item analysis, test equating, and optimal test construction
Reliability Analysis and Test equating using the MIRT package.
Albert, J.H. (1992). Bayesian estimation of normal ogive item response functions using
Gibbs sampling. Journal of Educational Statistics, 17, 251-269.
Brennan, R. L. (1992). Generalizability theory. NCME Instructional Module 14. Public
Domain.
Glas, C. A. W. (2012). Psychometric theory. Reader, University of Twente.
Glas, C. A. W. (2010). The MIRT package. Software, Reader, University of Twente.
Holman, R., & Glas, C. A. W. (2005). Modeling non-ignorable missing data mechanisms
with item response theory models. British Journal of Mathematical and Statistical
Psychology. 58, 1-17.
Korobko, O.B., Glas, C.A.W., Bosker, R.J. & Luyten, J.W. (2008). Comparing the difficulty
of examination subjects with item response theory. Journal of Educational
Measurement, 45, 139–157.
Kim, S.-H. (2001). An evaluation of a Markov Chain Monte Carlo method for the Rasch
model. Applied Psychological Measurement, 25, 163-176.
Lecture 2
Subjects
Workshop 2
Literature
Lecture 3
Subjects
Workshop 3
Literature
Lecture 4
Subjects
Practice 4
Literature
Item Response Theory
Introduction to maximum likelihood estimation and testing methods. Applications of IRT:
adaptive testing, analysis of item bias (differential item functioning, DIF) and handeling
of missing data.
DIF analysis using the MIRT package.
As for Lecture 1.
Introduction to Bayesian Item Response Modeling
Introduction to Bayesian item response modeling with applications: the testlet model,
models with mutiple raters, models for missing data, among other things.
Bayesian software using OpenBugs and R-packages.
Kim, S.-H. (2001). An evaluation of a Markov Chain Monte Carlo method for the Rasch
model. Applied Psychological Measurement, 25, 163-176.
Albert, J.H. (1992). Bayesian estimation of normal ogive item response functions using
Gibbs sampling. Journal of Educational Statistics, 17, 251-269.
Multilevel IRT modeling
Introduction to general multilevel models and multilevel IRT models, working with
plausible values.
Bayesian software using OpenBugs and R-packages.
As for Lecture 3.
Course location
Twente University, room:
16 December 2014
Ravelijn 2334
17 December 2014
Ravelijn 2334
18 December 2014
Spiegel 7
19 December 2014
Ravelijn 2336
Detailed information will be sent to the participants shortly before the start of the course.