Handbook of Item Response Theory, Volume Three: Applications

Abstract

Item response theory (IRT) has become the scaffolding for test development in many large-scale assessment programs. In addition to its main development base in educational and psychological measurement, IRT modeling has been applied in more social and behavioral disciplines such as health measurement, medical research, marketing research, political science, survey, and growth measures. Over the last several decades, the development in psychometric analyses based on IRT modeling has advanced dramatically in solving routine issues in the whole process of test development. Many innovative approaches and models for psychometric analyses have been proposed. In addition to the basic IRT models for dichotomous and polytomous item responses, multidimensional, multilevel, and explanatory IRT models have been developed. Further, nonparametric IRT models and models for nonmonotone items have been developed to meet different measurement requirements. On the other hand, due to the increase of online testing, models for process data such as item response time have been proposed to better understand the responding process in test-taking. All these standard and extended IRT models have been introduced and elaborated in Handbook of Item Response Theory, Volume one: Models (van der Linden, 2018a). The guidance related to how to apply these advances in psychometric modeling to enhance the practices is needed by both practitioners and researchers. Handbook of Item Response Theory, Volume Three: Applications by van der Linden (2018c) is such a reference book, providing a comprehensive perspective of applying the advances in IRT modeling to solve psychometric issues and challenges. The book consists of five sections. Each section covers topics related to different practical aspects of applications of IRT in operational settings: item calibration, person scoring, test design and computer programs for item response analysis. In addition, some applications of IRT models in measuring latent construct in different disciplines are highlighted in one section. Each chapter follows about the same structure in presenting the information, often starts with an introduction to the topic, continues with technical details and ends with conclusions and discussions which summarizes what has been included in the chapter and what are potential explorations in the future. The first section addresses item calibration and analysis. The five chapters in this section are related to item calibration, linking, dimensionality assessment, and differential item functioning analysis. Chapters 1 and 5 both focus on item calibration. Chapter 1 highlights the importance of calibration designs for optimality. Examples related to optimal calibration designs for the two-parameter IRT models and polytomous IRT models are presented. Calibration design is a factor that does not often draw attention to test developers in item calibration in practice. However, the significance of this factor in achieving optimality should be considered in item calibration. This chapter provides detailed information about the impact of calibration design on item calibration. Chapter 5, on the other hand, elaborates on the issues in calibration of technology-enhanced (TE) items. The use of unidimensional calibration of TE items is illustrated. The issues related to dimensionality and data-model fit are discussed. Chapter 2 introduces an emerging approach for linking with the consideration of both estimation and linking errors. Though more in-depth comparison between the current linking approaches and the presented one is needed, the issues raised related to the current standard practices warrant further explorations. Chapter 3 provides a comprehensive review of dimensionality assessment. Issues and recommendations are provided for practitioners. Differential item functioning (DIF) analysis is a routine procedure in item quality evaluation. Chapter 4 focuses on DIF MEASUREMENT: INTERDISCIPLINARY RESEARCH AND PERSPECTIVES 2021, VOL. 19, NO. 3, 186–189 https://doi.org/10.1080/15366367.2021.1936390