This Research Commentary addresses the need for an instrument abstract???termed an Interpretation and Use Statement (IUS)???to be included when mathematics educators present instruments for use by others in journal articles and other communication venues (e.g., websites and administration manuals). We begin with presenting the need for IUSs, including the importance of a focus on interpretation and use. We then propose a set of elements???identified by a group of mathematics education researchers, instrument developers, and psychometricians???to be included in the IUS. We describe the development process, the recommended elements for inclusion, and two example IUSs. Last, we present why IUSs have the potential to benefit end users and the field of mathematics education.

This article investigates the implementation of inquiry-oriented instruction in 20 undergraduate mathematics classrooms. In contrast to conventional wisdom that active learning is good for all students, we found gendered performance differences between women and men in the inquiry classes that were not present in a noninquiry comparison sample. Through a secondary analysis of classroom videos, we linked these performance inequities to differences in women's participation rates across classes. Thus, we provide empirical evidence that simply implementing active learning is insufficient, and that the nature of inquiry-oriented classrooms is highly consequential for improving gender equity in mathematics.

This study tested how different implementations of explaining and drawing activities affect learning from a multimedia science lesson. After studying a multimedia slideshow about the human respiratory system, college students (n = 145) were assigned randomly to one of four learning activity conditions: write explanations before drawing pictures (explain-then-draw group), draw pictures before writing explanations (draw-then-explain group), study provided explanations before drawing pictures (provided explanation-then-draw), or study provided pictures before writing explanations (provided drawing-then-explain). One week following the learning activity, all students completed post-tests of their understanding. Results from the learning activity supported the scaffolding hypothesis: students generated better explanations when they used provided (rather than their own) drawings, and they generated better quality drawings when they used provided (rather than their own) explanations. However, this difference in learning activity performance did not correspond to higher performance on the delayed post-tests. We discuss implications for how to best sequence and scaffold generative activities.

In several author name disambiguation studies, some ethnic name groups such as East Asian names are reported to be more difficult to disambiguate than others. This implies that disambiguation approaches might be improved if ethnic name groups are distinguished before disambiguation. We explore the potential of ethnic name partitioning by comparing performance of four machine learning algorithms trained and tested on the entire data or specifically on individual name groups. Results show that ethnicity-based name partitioning can substantially improve disambiguation performance because the individual models are better suited for their respective name group. The improvements occur across all ethnic name groups with different magnitudes. Performance gains in predicting matched name pairs outweigh losses in predicting nonmatched pairs. Feature (e.g., coauthor name) similarities of name pairs vary across ethnic name groups. Such differences may enable the development of ethnicity-specific feature weights to improve prediction for specific ethic name categories. These findings are observed for three labeled data with a natural distribution of problem sizes as well as one in which all ethnic name groups are controlled for the same sizes of ambiguous names. This study is expected to motive scholars to group author names based on ethnicity prior to disambiguation.

Assessment developers are increasingly using the developing technology of machine learning in transforming how to assess students in their science learning. I argue that these algorithmic models further embed the structures of inequality that are pervasive in the development of science assessments in how they legitimize certain language practices that protect the hierarchical standing of status quo interests. My argument is situated within the broader emerging ethical challenges around this new technology. I apply a raciolinguistic equity analysis framework in critiquing the new black box that reinforces structural forms of discrimination against the linguistic repertoires of racially marginalized student populations. The article ends with me sharing a set of tactical shifts that can be deployed to form a more equitable and socially-just field of machine learning enhanced science assessments.

Research exploring students' learning from physical and virtual labs has suggested that on the whole, students learn science content just as well, if not better from virtual labs as they do from physical labs. However, the affordances of physical labs might support the learning of specific skills and competencies that are just as crucial for learning science. In this study, we examined students' discussions as they worked on physical and virtual labs to better understand how they learned from each, and the kinds of learning that each type of lab supported. One hundred and fifteen 6th grade students from three science teachers' classes participated in this study. We examined audio data from all available groups as they engaged in physical and virtual labs (n =14 groups; physical,n= 8 groups; virtual,n= 6 groups). We found that students conducting physical labs engaged in a significantly higher proportion of talk related to setting up apparatus and taking measurements and calculating outputs. Students who performed virtual labs, on the other hand, engaged in significantly more discussions about making predictions and understanding patterns of relationships between variables, and interpreting science phenomena. While students in the Virtual condition engaged in discussions that were more focused on the relationships between science ideas, students in the Physical condition learned science practices related to planning and carrying out investigations that are equally valuable. Our findings suggest that learning from one experimental modality may complement and supplement the relative weaknesses of the other, indicating a need for strategically combining the two. Implications and future directions are discussed.

Constructing causal mechanistic explanations of observable phenomena is a key science practice that is often challenging for students as most mechanisms involve interactions of unobservable entities and activities. In this study, we examined how gesturing with a computer simulation that depicts the molecular mechanism of thermal conduction supported middle-school students in constructing causal mechanistic explanations. We designed a gesture-augmented computer simulation in which students were cued to use hand gestures to control the simulation. These cued gestures represent core causal interactions of conduction and they prompt students to physically engage with the simulation in conceptually meaningful ways. In this study, we examined how 21 students used the simulation and explained thermal conduction in a semi-structured interview, followed by a mixed-methods analysis. Quantitative analysis shows that students moved toward articulating the canonical causal mechanistic explanation of thermal conduction using the simulation. Three representative cases were identified to explore how students' explanations were facilitated by cued gestures. The analysis shows two main ways the cued gestures supported all students in the study: (a) by helping them attribute causal agency to molecules rather than an entity called Heat, and (b) by reifying the core mechanism of molecular collisions in conduction. Furthermore, the case studies show how each student's unique ways of sensemaking impacted their gesture use. Implications for instruction with gestures and design of augmented environments are discussed.

The micro-level analyses of how students’ self-regulated learning (SRL) behaviors unfold over time provides a valuable framework for understanding their learning processes as they interact with computer-based learning environments. In this paper, we use log trace data to investigate how students self-regulate their learning in the Betty’s Brain environment, where they engage in three categories of open-ended problem-solving actions: information seeking, solution construction and solution assessment. We use Epistemic Network Analysis (ENA) to provide us with an overall understanding of the co-occurrences between action types both within and between the three action categories. Comparisons of epistemic networks generated for two groups of students, those with low and high performance, provided us with insights into their self-regulated behaviors. © 2021, Springer Nature Switzerland AG.

Self-regulated learning (SRL) with advanced learning technologies has shown to significantly augment learners' performance across contexts. Yet studies find learners lack sufficient SRL skills to successfully implement strategies (e.g., judgments of learning, note taking, self-testing, etc.). Current research does not fully explain how and why this failure of effective strategy deployment occurs. We used principle component analysis (PCA) on process data (i.e., log files) from 190 undergraduates learning with MetaTutor, a hypermedia-based intelligent tutoring system, to explore underlying patterns in the frequency of strategy deployment occurring with and without pedagogical agent scaffolding to better understand any underlying structures of system- and learner-initiated cognitive and metacognitive SRL strategy use. Results showed that the system's underlying architecture deploys processes corresponding to both the phases of learning and type of effort allocation according to Winne's (2018) Information Processing Theory of SRL. However, learner-initiated processes for those who received scaffolding only displayed strategy deployment that corresponded to the type of effort allocation required of the processes (i.e., more effortful constructionist processes like note-taking versus short canned responses for judgements of learning). Additionally, results suggest all learners deploy strategies based on the familiarity of processes. Regression models using these principle components outperformed raw frequency models for capturing post-test learning performance across all participants.