A5029351368- Student Assessment and Feedback
- Online Learning and Analytics
- Educational Assessment and Pedagogy
- Educational Technology and Assessment
- Innovative Teaching and Learning Methods
- Innovative Teaching Methods
- Behavioral Health and Interventions
- Science Education and Pedagogy
- School Choice and Performance
- Sports Analytics and Performance
- Complex Systems and Decision Making
- Manufacturing Process and Optimization
- Diverse Music Education Insights
- Intelligent Tutoring Systems and Adaptive Learning
- Early Childhood Education and Development
- Mental Health Research Topics
- Neural and Behavioral Psychology Studies
- Art Education and Development
- Reflective Practices in Education
- Neural Networks and Applications
- Education and Critical Thinking Development
- Artificial Intelligence in Law
- Business Process Modeling and Analysis
- Advanced Statistical Modeling Techniques
- Educational Research and Pedagogy
American Institutes for Research
2017-2024
Educational Testing Service
2022
Center for Assessment
2017
University of Colorado Boulder
2016
Boğaziçi University
2010
This mini review summarizes the current state of knowledge about automatic item generation in context educational assessment and discusses key points pipeline. Assessment is critical all learning systems digitalized assessments have shown significant growth over last decade. leads to an urgent need generate more items a fast efficient manner. Continuous improvements computational power advancements methodological approaches, specifically field natural language processing, provide new...
This study explored the effectiveness of extended time (ET) accommodations in 2017 NAEP Grade 8 Mathematics assessment to enhance educational equity. Analyzing process data through an XGBoost model, we examined if early interactions with items could predict students’ likelihood requiring ET by identifying those who received a timeout message. The findings revealed that 72% students disabilities (SWDs) granted did not use it fully, while about 24% lacking were still actively engaged when...
The speed–accuracy trade-off (SAT) suggests that time constraints reduce response accuracy. Its relevance in observational settings—where (RT) may not be constrained but respondent speed still vary—is unclear. Using 29 data sets containing from cognitive tasks, we use a flexible method for identification of the SAT (which test extensive simulation studies) to probe whether holds. We find inconsistent relationships between and accuracy; marginal increases an individual do necessarily predict...
Chatbots have been an interesting application of natural language generation since its inception. With novel transformer based Generative AI methods, building chatbots become trivial. which are targeted at specific domains such as medicine, psychology, and general information retrieval implemented rapidly. This, however, should not distract from the need to evaluate chatbot responses. Especially because community does entirely agree upon how effectively applications. this work we discuss...
This article describes a 4-year study of experienced high school biology teachers' participation in five-step professional development experience which they iteratively studied student ideas with the support set learning progressions, designed formative assessment activities, practiced using those activities their students, enacted and then reflected on next steps to guide instruction. Drawing classroom artifacts responses pre–post assessment, we examined alignment teacher-created tasks as...
Artificial Neural Networks (ANNs) have been proposed as a promising approach for the classification of students into different levels psychological attribute hierarchy. Unfortunately, because such classifications typically rely upon internally produced item response patterns that not externally validated, instability ANN estimates probabilities may be widely appreciated. The present study illustrates problem with both empirical and simulated data. In particular, it is shown when an "trained"...
Abstract The purpose of this study was to explore high school course‐taking sequences and their relationship college enrollment. Specifically, we implemented sequence analysis discover common trajectories in math, science, English language arts using transcript data from a recent nationally representative survey. Through clustering, reduced the complexity examined sequences. Classification tree, random forests, multinomial logistic regression analyses were used between course students...
Abstract Research shows that the intensity of high school course‐taking is related to postsecondary outcomes. However, there are various approaches measuring students’ course‐taking. This study presents new measures coursework rely on differing levels quantity and quality coursework. We used these indices provide a current description variations in across grades student subgroups using nationally representative dataset, High School Longitudinal Study 2009. Results showed for emphasizing gaps...
The purpose of this study was to display the effectiveness restoring test items on relevancy responses construct be measured. 22 Items were selected from Turkish version PISA 2006 science component. Based revisions made revised formed (PISA-RT). PISA-RT and original tests (PISA-OT) administered two independent group 30 students in each. These equivalent groups. who took performed significantly better than one PISA-OT all test.
In this study, we explore the application of process mining techniques on assessment log data to problem-solving strategies in Algebra. By analyzing sequences student activities, demonstrate significant potential identifying that lead successful and unsuccessful outcomes. Our findings reveal students who successfully solve problem tend follow one three structured strategies, displaying a systematic filling boxes Pascal's triangle. Conversely, those falter often start with correct strategy...
The speed-accuracy tradeoff suggests that responses generated under time constraints will be less accurate. While it has undergone extensive experimental verification, is clear whether applies in settings where pressures are not being experimentally manipulated (but respondents still vary their utilization of time). Using a large corpus 29 response datasets containing data from cognitive tasks without manipulation pressure, we probe the holds across variety using idiosyncratic within-person...