Item analysis (IA) is commonly used to describe difficulty and discrimination indices of multiple
true-false (MTF) questions. However, item analysis is basically a plain descriptive analysis with
limited statistical value. Item response theory (IRT) can provide a better insight into the difficulty
and discriminating ability of questions in a test. IRT consists of a collection of statistical models that
allows evaluation of test items (questions) and test takers (examinees) at the same time. Specifically,
this article focuses on two-parameter logistic IRT (2-PL IRT) model that is concerned with estimation
of difficulty and discrimination parameters. This article shows how 2-PL IRT analysis is performed in
R software environment, guides the interpretation of the IRT results and compares the results to IA on
a sample of MTF questions.