Past perspectives and new opportunities for the explanatory item response model

Abstract

Models of word reading that simultaneously take into account item-level and person-level fixed and random effects are broadly known as explanatory item response models (EIRM). Although many variants of the EIRM are available, the field has generally focused on the doubly explanatory model for modeling individual differences. Moreover, the historical application of the EIRM has been a Rasch version of the model where the item discrimination values are fixed at 1.0 and the random or fixed item effects only pertain to the item difficulties. The statistical literature has advanced to allow for more robust testing of observed or latent outcomes, as well as more flexible parameterizations of the EIRM. The purpose of the present study was to compare four types of Rasch-based EIRMs (i.e., doubly descriptive, person explanatory, item explanatory, doubly explanatory) and more broadly compare Rasch and 2PL EIRM when including person-level and item-level predictors. Results showed that not only was the error variance smaller in the unconditional 2PL EIRM compared to the Rasch EIRM due to including the item discrimination random effect, but that patterns of unique item-level explanatory variables difference between the two approaches. Results are interpreted within the context of what each statistical model affords to the opportunity for describing and explaining differences in word-level performance.

Differential nonlinear relations of language proficiencies to reading and math achievement in Spanish or English.

“…In addition, more research on item-level characteristics using item-response theory (IRT) would help identify which types of items and their features vary across levels of ELP. Methods such as differential item functioning (Buono & Jang, 2021) or explanatory IRT (Petscher et al, 2020) could be used to detect item-level differences. These types of questions would advance the current literature by showing not if ELP and language of assessment interact but at what level (e.g., item, person) and with what item characteristics (e.g., complex word problems vs. calculation).…”

Analyzing the Effects of Test, Student, and School Predictors on Science Achievement: An Explanatory IRT Modeling Approach

“…EIRM is a framework that allows measuring covariates in item sets, student groups, or interactions between item sets and student groups (De Boeck & Wilson, 2004). The main function of EIRM is (1) the odds of accuracy at the item level, (2) individual differences in item-level accuracy, (3) how much of the variance of item level accuracy is due to differences between items rather than interpersonal differences, (4) the chosen person, item predictors and interactions to reveal the explanation of variances (Petscher et al, 2020). Rather than calculating the descriptive effects on the student's feature level or item difficulty, EIRM allows obtaining information by taking into account the explanatory variables with the responses.…”

Modeling Item-Level Heterogeneous Treatment Effects With the Explanatory Item Response Model: Leveraging Large-Scale Online Assessments to Pinpoint the Impact of Educational Interventions

“…While the data generating process of this simulation was based on a 1PL model, a 1PL approach may not be appropriate for educational assessments in which items vary in their discriminations as well as their difficulties. Advances in estimation methods such as profile-likelihood (Jeon & Rabe-Hesketh, 2012) have enabled exploration of the 2PL EIRM that models item discriminations as either fixed quantities to be estimated, as in the mirt (Chalmers, 2012) or PLmixed (Rockwood & Jeon, 2018) R packages and the gllamm Stata program (Skrondal & Rabe-Hesketh, 2004), or as random variables to themselves be explained by the predictors in both frequentist (Cho et al, 2014; Petscher et al, 2020, using Mplus) and Bayesian paradigms (Bürkner, 2019, using R’s brms). However, sensitivity analyses based on a 2PL data generating process presented in the OSM demonstrate that varying item discriminations have no impact on treatment effect bias, false positive rates, or statistical power but do result in slightly less accurate SE s (though still within approximately 5% of their true values).…”