W. Holmes Finch

Various manifestations of Artificial Intelligence (AI), including Generative AI (GenAI), are being developed and deployed in higher education. The embrace of AI tools, especially those targeted at students, has created questions surrounding the future of teaching and learning. While there is much excitement and hope about AI’s pedagogical potential, the question of how GenAI reshapes assignment completion is not simply one issue among many, but one that raises a fundamental concern with academic integrity. In such context, scholars and policy makers must consider student perspectives, especially ones regarding the way students complete the everyday assignments that provide structure to university life. The AI age necessitates not only a canvasing of general attitudes, but a granular investigation of student perception of peer use of the technology. This study centers on two broad questions, including R1: Whether students think AI encourages dishonesty and, R2: If they perceive that it gives users an unfair advantage. To gain some insight into these questions, current undergraduate students (N = 214) at two public North American universities responded to our exploratory survey. Our research uncovers strong feelings towards AI technology in higher education, which underscores the importance of including all stakeholders—especially undergraduate students—in the broader conversations about the fate of pedagogical achievement and academic integrity in a rapidly changing educational environment.

Interrater reliability is a critical aspect of measurement quality, particularly in assessments that rely on subjective judgment. However, interrater reliability estimates vary, and such variability can introduce bias or reduce the accuracy of observed scores, especially when comparing across groups or conditions. Understanding and accounting for these differences is essential when interpreting reliability in applied settings such as education, psychology, and performance evaluation. This study addresses the need for more nuanced approaches to evaluating interrater reliability across groups. Specifically, in this study, we examine generalizability theory (GT) and structural equation modeling (SEM) that enable direct testing of differences in reliability coefficients across groups. A simulation study compared a proposed method grounded in GT and SEM to the W statistic for reliability coefficient comparisons. Results demonstrate that the proposed method consistently outperforms the W statistic in terms of both Type I error control and statistical power, particularly when sample sizes are moderate to large or when variance in rater agreement exists across groups. These findings underscore the importance of explicitly modeling differences in interrater reliability and provide researchers with a more robust tool for evaluating the consistency of ratings across diverse contexts and populations.

This seminar describes estimation techniques for item response theory models when the latent trait is not normally distributed. These alternatives to the standard approach, which assumes a normally distributed latent trait, provide more accurate estimates with skewed, bimodal, and zero-inflated latent traits. It is designed for researchers who are faced with such skewed constructs.

This workshop covers application and theory for fitting latent variable mixture models, with a focus on specification, estimation, model selection, and robust inference. Application to real world data analysis is emphasized with sufficient conceptual foundation to understand the results and support research publication.

Mathematics anxiety and statistics anxiety, along with low self‐efficacy, often negatively impact student performance and interest in STEM fields. This research examines the unique patterns and interrelationships affecting students' confidence and dispositions in mathematical and statistical domains. Despite extensive research on individual effects and mathematics dispositions, limited studies examine these variables' concurrent mediation analyses and direct and indirect effects. The mediating role of self‐efficacy in the relationship between mathematics anxiety and statistics anxiety remains underexplored. University students enrolled in introductory statistics courses completed self‐report surveys on their mathematics dispositions and demographic information to investigate individual differences related to mathematics anxiety, statistics anxiety, and self‐efficacy. Using structural equation modeling, we tested (a) the direct association between mathematics anxiety and statistics anxiety, (b) the indirect pathways from mathematics anxiety to statistics anxiety through mathematics self‐efficacy, and (c) the indirect pathway through statistics self‐efficacy. All variables were significantly correlated. Direct associations were significant, except for mathematics self‐efficacy and statistics anxiety. The significant mediating role of mathematics self‐efficacy on mathematics anxiety and statistics anxiety indicates the importance of self‐efficacy in dispositional factors. Understanding these connections allows educators to develop targeted interventions to boost self‐efficacy, thereby reducing anxiety and potentially improving student performance and interest in STEM fields.

Applied and simulation studies document model convergence and accuracy issues in differential item functioning detection with multilevel models, hindering detection. This study aimed to evaluate the effectiveness of various estimation techniques in addressing these issues and ensure robust DIF detection. We conducted a simulation study to investigate the performance of multilevel logistic regression models with predictors at level 2 across different estimation procedures, including maximum likelihood estimation (MLE), Bayesian estimation, and generalized estimating equations (GEE). The simulation results demonstrated that all maintained control over the Type I error rate across conditions. In most cases, GEE had comparable or higher power compared to MLE for identifying DIF, with Bayes having the lowest power. When potentially important covariates at levels-1 and 2 were included in the model, power for all methods was higher. These results suggest that in many cases where multilevel logistic regression is used for DIF detection, GEE offers a viable option for researchers and that including important contextual variables at all levels of the data is desirable. Implications for practice are discussed.

Exploratory factor analysis (EFA) is a widely used tool in the social sciences. Researchers employ it to identify the latent structure underlying observed indicator variables during the process of scale development, theory construction, and comparison of various constructs. One of the most important aspects of conducting EFA is determining the number of factors to retain. There exist a number of techniques for this purpose, but none have been identified as uniformly optimal in all situations. The purpose of this simulation study is to compare several such techniques in the context of dichotomous and ordinal indicator variables (corresponding to items on an instrument). Some of the methods investigated in this study include well-established techniques, such as parallel analysis and the minimum average partial correlation, as well as newly developed ones, such as out-of-sample prediction error and the next eigenvalue sufficiency test. The results of the study demonstrate that a Bayesian estimation approach and the out-of-sample prediction error method are particularly effective for identifying the number of factors to retain. The implications for practice are discussed.

This seminar provides researchers with the theoretical framework and hands-on implementation needed to specify, estimate, and interpret complex nonlinear relationships among latent variables in R and Mplus. The nonlinear effects include latent variable interactions, quadratic relationships among latent variables, nonlinear growth, and splines for latent variable models. Participants will learn the statistical frameworks underlying these models and will be provided with reproducible software scripts. The seminar concludes with advice on how to effectively present the results of nonlinear latent variable models. The focus throughout is on application and interpretation of the models.