A Mixed Methods Design for Assessing Physics Learning in the Online Learning Environment

Abstract

This study explored a Bayesian assessment model for physics students in motion learning. The simulated data was applied in examination of the Bayesian assessment model, The study used a mixed-methods design. The exploratory sequential model was developed based on a motion learning student model, which was a structured data collection template. The combination of the student model and the Bayesian network model provided an assessment tool for assessing physics students’ learning in a dynamic process. The study reported that there were three different patterns for a physics student motion learning: lower performance, middle performance, and higher performance. In each pattern, the students may have different performance combinations of the twelve bottom components. These are shown in Figure 4 and used to collect students’ performance data.

Diagnostically Cognitive Assessment of Complex Semantic Explanations and Learning: A Bayesian Network Approach

“…These models can be used to develop techniques for coding analysis and scoring of such records or procedures of performance (Hollnagel, 2003). In principle, both item response and Bayesian evidence models (Zhang, 2022) can be applied to evidence variables based on such coding of a performance. Such models can support inferences about overall proficiency, and diagnostic inferences about specific knowledge, reasoning, and procedures that are involved in and underlie cognitive expertise in a domain (Alexander, 2003)…”

“…In this paper, we report the results of the investigation of a BN approach to diagnostically cognitive assessment that we believe can contribute to meeting there are challenges (Zhang, 2022). We applied Mislevy's evidence-centered assessment design (ECD) framework to develop a diagnostic cognitive assessment model in which Bayesian probability network (BPN models) were developed and used to make inferences about the components of students' knowledge and proficiency in a domain for problem-solving in intermediate statistics (Mislevy, Steinberg, Almond, Breyer, & Johnson 2001;Zhang, 2007;Zieky, 2014).…”

“…Bayesian networks (BNs) are known as belief networks, which are represented in a directed acyclic graph to model an assessment model (Zhang, 2022). Kwan, Chow, Law and Lai (2008) describe the Bayesian network:…”

A Mixed Assessment for the Science Learning via a Bayesian Network Representation

“…The Bayesian network represents the semantic/qualitative components and constructs the probability model separately (Koller & Friedman, 2009). Thus, the combination of a student model and a Bayesian network model provides a path to develop an assessment framework used to collect and represent the data (Zhang, 2022).…”

“…This framework consists of three critical components: declarative, procedural, and strategic (conditional) knowledge components. Three components supported the framework of the student model, and are further connected to the Bayesian network model ( (Mislevy, Almond & Lukas, 2004;Zhang, 2022). Thus, a structured assessment model was developed and students' learning process, progress, and mastery can be recognized.…”