A Mixture IRTree Model for Extreme Response Style: Accounting for Response Process Uncertainty

Abstract

This paper presents a mixture item response tree (IRTree) model for extreme response style. Unlike traditional applications of single IRTree models, a mixture approach provides a way of representing the mixture of respondents following different underlying response processes (between individuals), as well as the uncertainty present at the individual level (within an individual). Simulation analyses reveal the potential of the mixture approach in identifying subgroups of respondents exhibiting response behavior reflective of different underlying response processes. Application to real data from the Students Like Learning Mathematics (SLM) scale of Trends in International Mathematics and Science Study (TIMSS) 2015 demonstrates the superior comparative fit of the mixture representation, as well as the consequences of applying the mixture on the estimation of content and response style traits. We argue that methodology applied to investigate response styles should attend to the inherent uncertainty of response style influence due to the likely influence of both response styles and the content trait on the selection of extreme response categories.

Dynamic Response Strategies: Accounting for Response Process Heterogeneity in IRTree Decision Nodes

“…There is a wide range of psychometric approaches accounting for heterogeneity in response processes with regard to RS, whereby the distinction between trait-based and RS-based processes has mainly been considered on the between-person level. For instance, mixture Rasch models (e.g., Austin et al, 2006;Gollwitzer et al, 2005, Meiser & Machunsky, 2008, mixture IRTree models (e.g., Khorramdel et al, 2019, Kim & Bolt, 2021, and a general mixture IRT model (Tijmstra et al, 2018) were proposed, which all can be used to identify latent classes of respondents who provide item responses based on different processes, such as responses influenced by response styles or not (i.e., solely traitbased responses). A limitation of such models is that the response process heterogeneity is strictly related to between-person effects so that possible class switches cannot be detected.…”

“…2. Thereby, the process loadings of the DRSM are defined to be nonnegative across all items, which is a frequently made assumption in IRT modeling (e.g., Jin & Wang, 2014;Kim & Bolt, 2021;Meiser et al, 2019). We consider this a reasonable constraint also for the loading trajectories, since variations in test-taking effort should result in a varying degree of trait and RS involvement, that is, in a varying size of the loadings, whereas a change toward negative loadings would rather imply a qualitatively different effect of such latent personal characteristics on response selection (e.g., high trait levels would then be associated with low instead of high response categories).…”

“…Various item response theory (IRT) approaches have been proposed that facilitate such control of RS effects under conditions in which the underlying response processes are homogeneous across persons and over items, thus assuming a stable response strategy over the course of a questionnaire (e.g., Böckenholt 2017;Bolt & Newton, 2011;Henninger & Meiser, 2020;Plieninger & Meiser, 2014;Wetzel & Carstensen, 2017). Extensions of such models can further account for some kind of heterogeneity of response processes over discrete conditions, either with a focus on latent classes of respondents (e.g., Kim & Bolt, 2021;Tijmstra et al, 2018;von Davier & Yamamoto, 2007) or with a focus on within-person changes across measurement occasions (e.g., Ames & Leventhal, 2021;Colombi et al, 2021;Weijters et al, 2010). Other approaches include unsystematic item-by-item fluctuations of response strategies within persons (e.g., Plieninger & Heck, 2018;Tijmstra & Bolsinova, in press;Ulitzsch et al, 2022).…”

Modeling Rapid Guessing Behaviors in Computer-Based Testlet Items

“…As the number of testlets increases, the TM-RT becomes increasingly complex, as shown in Equations (3), (4), (5). According to the experiences of many past studies (e.g., Huang, 2020; Kim & Bolt, 2021; Man et al, 2019), using the following semi-informative priors enables efficient estimation of the TM-RT parameters via Bayesian methods with Markov chain Monte Carlo (MCMC): b j ∼ N (0, 0.1),β j ∼ N (0, 0.1),α j ∼ N (0, 0.1).…”

Mixture item response tree model for measuring extreme response style in multiple psychological scales

“…최근 선행연구에서는 IRTree 모형을 혼합 문항반응모형으로 확장한 혼합문항반응나무 모형(Mixture item response tree model: 이하 MixIRTree model)이 제안되었다 (Khorramdel, von Davier, & Pokropek, 2019;Kim & Bolt, 2021;Tijmstra, Bolsinova, & Jeon, 2018) (Paulhus, 1991), 최근 연구에서는 일부 응답자 에게는 ERS가 척도의 내용에 영향을 받는 특 성(content-relevant)으로 보고되고 있다 (Kim & Bolt, 2021;Meiser, Plieninger, & Henninger, 2019). 예를 들면, 어떤 응답자들은 높은 잠재 목표특성치 때문에 '매우 그렇다'를 선택할 수 있다 (Kim & Bolt, 2021). 반면 어떤 응답자 들이 문항에 빠르게 응답하거나 응답 시 단순 하게 사고하는 특성 (Naemi, Beal, & Payne, 2009) (Bolt & Newton, 2011;Jackson & Messick, 1958;Paulhus, 1991;Wetzel et al, 2013).…”