In psychological research, it is often needed to measure latent person characteristics such as mathematics ability, anxiety, and attitude towards political issues. Tests and test scores play an important role. Traditionally, test scores were analyzed by classical test theory, but this has its limitations. Over the last decades, modern test theory has gained influence, since it allows us to analyze the data on the item level and to develop and test a measurement scale of item responses (Item Response Theory models or IRT).
This course offers a theoretical and practical introduction to IRT models. IRT principles are contrasted with CTT principles. Several IRT models are discussed: those for dichotomous item responses (e.g., right/wrong; agree/disagree) and those for categorical responses (e.g., wrong/partially correct/completely correct; multiple choice options). In each of these models, the probability of answering a particular response category is related to the latent characteristic or the latent trait, in a nonlinear way. Several applications of IRT are discussed, including how to calibrate a measurement scale, how to obtain person estimates and how to interpret them, how to find out whether all items ‘work’ the same way for different groups of test takers (differential item functioning), how to compare scores measured on similar but different tests (test equating), and how to make use of computers to make tests shorter (computerized adaptive testing). Several analyses will be practiced in R (with the predefined ltm package).
Coordinator
Prof.dr. W.J. Heiser
Room 3B19
Tel.: +31 (0)71 527 3828
E-Mail: heiser@fsw.leidenuniv.nl
Aims
Students
Acquire knowledge of and insight into the principles of relevant modern test theoretical models and techniques
Acquire knowledge of, insight into, and practical competence w.r.t. analyzing test data with IRT models
Acquire knowledge of and insight in several issues in applications of IRT to different fields of psychology
Literature
Embretson, S. E. & Reise, S. P. (2000). Item Response Theory for Psychologists. Mahwah, NJ: Lawrence Erlbaum Associates, Inc. ISBN10: 0805828192. Approximately €40 – €50.
Rizopoulos, D. (2006). ltm: An R package for latent variable modeling and Item Response Theory analyses. Journal of Statistical Software, 17(5), 1–25. Can be downloaded from http://www.jstatsoft.org/v17/a05/paper
Examination
Graded assignments during course and 1 final assignment.
From January 1, 2006 the Faculty of Social Sciences has instituted the Ephorus system to be used by instructors for the systematic detection of plagiarism in students’ written work. Please see the Additional Rules and Regulations, section 6.
Education method(s)
Lectures and software demonstrations.
Enrolment
Introduction and enrolment for courses of the first semester will take place August 27th 2009. Introduction and enrolment for courses of the second semester will take place in January 2010. More information will be available at the website of the Institute of Psychology.
Blackboard
Information available on www.blackboard.leidenuniv.nl