Developing latent constructs of dialogic interaction to examine the epistemic climate: Rasch modeling

Abstract

This study investigates two latent constructs (Engagement and Value) of dialogic interaction to examine the epistemic climate. Since the new reform movement emphasizes creating generative learning environments, it is important to examine whether a classroom promotes students’ knowledge generation or limit students’ epistemic growth through rote memorization. At this point, it is inevitable to focus on dialogic interaction (one of the epistemic practices), because how students engage with and view dialogic interaction provides meaningful information about the epistemic climate. By employing Rasch modeling, this study tested a statistical validity of two latent constructs of dialogic interaction. The findings in this study highlight that the two latent constructs are theoretically and statistically valid and can be used to gauge the epistemic climate through students’ engagement with and value of dialogic interaction. While understanding the epistemic climate cannot solely occur through a single perspective, this study plays a significant role in adding a valid instrument. Such instruments help researchers comprehend the epistemic features of learning environments and provide a practical way to reflect students' lived experiences to learn about the overall atmosphere of the learning environment.

Does employee engagement mediate the nexus of job resource and employee turnover intentions?

“…The results of each scale were calculated by summing the responses to the items. Bae et al (2021) instrument refinement technique was applied in ensuring that items that adequately represent the presumed dimensions of a construct were included. Hair et al (2019) and Henseler and Schuberth (2020) recommended criteria for construct standards estimates were applied.…”

Validity aspects in measuring evolution acceptance: Evidence from surveys of preservice biology teachers and creationists

“…Multi‐dimensional Rasch models can minimize validity problems encountered when unidimensional models are applied to multifaceted instruments (e.g., multiple subscales derived from theory; Adams et al, 1997; Cheng et al, 2009). Thus, multi‐dimensional Rasch models are frequently applied in science education research (e.g., Bae et al, 2021; Ding et al, 2016), and in particular in evolution education research (Fiedler et al, 2019; Sbeglia & Nehm, 2019). To evaluate the dimensionality of each instrument, we conducted principal component analyses of Rasch residuals (PCAR) of a one‐dimensional model (i.e., instruments' item sets were treated as measuring one latent construct; Bond & Fox, 2015).…”