Application of nonparametric item response theory in determining the one-dimensionality and adaptability of TOEFL iBT listening test

Abstract

Abstract Listening comprehension in English, as one of the most fundamental skills, has an essential role in the process of learning English. Non-parametric item Response Theory (NIRT) is a probabilistic-nonparametric approach to item response theory (IRT) which determines the one-dimensionality and adaptability of test. NIRT techniques are a useful tool for researchers who wish to construct one-dimensional tests. The current study utilized the NIRT to examine adaptability of TEOFL iBT listening test, administered on 400 EFL university students in the Iranian context. The results illustrated no main concerns in terms of item adaptability. It was concluded that the ordering of items according to their mean is invariant across examinees. Dimensionality analysis results depicted that the test is one dimensional -confirming evidence of the validity of the test in measuring a single ability dimension. The empirical application depicted a potential and feasible approach, whereby NIRT could be used as a valuable method for exploring the behavior of scaled items in response to varying levels of a latent trait in education research.

Keywords

• Double monotonicity model
• Nonparametric item response theory
• Monoton homogeneity model
• Listening comprehension
• TOEFL iBT