An Application of Item Response Time: The Effort‐Moderated IRT Model

Abstract

The validity of inferences based on achievement test scores is dependent on the amount of effort that examinees put forth while taking the test. With low‐stakes tests, for which this problem is particularly prevalent, there is a consequent need for psychometric models that can take into account differing levels of examinee effort. This article introduces the effort‐moderated IRT model, which incorporates item response time into proficiency estimation and item parameter estimation. In two studies of the effort‐moderated model when rapid guessing (i.e., reflecting low examinee effort) was present, one based on real data and the other on simulated data, the effort‐moderated model performed better than the standard 3PL model. Specifically, it was found that the effort‐moderated model (a) showed better model fit, (b) yielded more accurate item parameter estimates, (c) more accurately estimated test information, and (d) yielded proficiency estimates with higher convergent validity.

Conditioning factors of test-taking engagement in PIAAC: an exploratory IRT modelling approach considering person and item characteristics

“…Solution behavior is not assumed to be known, but the switching-point from solution behavior to rapid guessing, which may differ across test-takers, is estimated as part of the model. In contrast, the effort-moderated IRT model proposed by Wise and DeMars (2006) incorporates a variable derived from response time to indicate solution behavior and whether or not the regular IRT model holds for a particular itemperson combination. To identify different response modes, Schnipke and Scrams (1997) proposed a log-normal mixture model of item response time assuming two types of response-time distributions, one for rapid guessing and the other for solution behavior (expressed as a bimodal empirical response time distribution).…”

“…For instance, the effort-moderated IRT model proposed by Wise and DeMars (2006) applies a 3-parameter logistic (PL) IRT model for responses given in the solution behavior mode, while a constant probability model is applied for rapid-guessing behavior. Information on disengagement can also be used to fine-tune the scoring of response behavior.…”

Prospective educators as consumers of empirical research: an authentic assessment approach to make their competencies visible

“…These effects could be intensified in the field of higher education, because of the lack of compulsory attendance (Wise and DeMars 2006;Wolf et al 2015). Students who organize their studies independently-what is explicitly desired-could tend to neglect low-stakes tests.…”

Enhancing a Computer-Based Testing Environment with Optimum Item Response Time

“…It is important to note that "The accuracy of such rapid guesses is typically at or near the chance level, as the responses are essentially random" (Lee & Jia, 2014, p. 2). Students' behaviors, such as rapid guessing, would influence the test results in various ways such as reliability and validity (Wise & Kong, 2005;Wise & DeMars, 2006). There are several studies that propose new approaches for improving test's measurements by focusing on response time (see Kong, Wise, Harmes, & Yang, 2006;Wise & DeMars 2006;Wise, Bhola, & Yang, 2006).…”