ABSTRACT
This chapter concentrate on nonparametric Item response theory (NIRT) models for the Practical Data analysis of polytomous item scores. A typical aspect of NIRT models is that they based on weaker assumptions than most parametric IRT models and, as a result, often fit empirical data better. The purposes of the chapter are twofold. First, three NIRT models for the analysis of polytomous item scores discussed, and several well-known IRT models, each being a special case of one of the NIRT models, mentioned. The NIRT models are the nonparametric partial credit model, the nonparametric sequential model, and the nonparametric graded response model. Then, the hierarchical relationships between these three NIRT models proved. Second, an overview of statistical methods available and accompanying software ordered latent class analysis of polytomous item scores from questionnaires provided. Also, the kind of information provided by the statistical methods, and how this information might be used for drawing conclusions about quality of measurement by questionnaires explained.
