Our inaugural DREAM Edu 2023 Workshop on November 2 was a smashing success! Current and aspiring researchers came together to dive into ongoing projects and explore potential collaborations within the CourseKata community. The event was highlighted by lightning talks from researchers at five different institutions, making it a truly memorable and impactful day.
We also launched our brand-new CourseKata researcher website! This site is designed to be the ultimate resource hub, empowering researchers and students working with CourseKata data. Here’s what it offers:
- Central Hub: All resources related to CourseKata research in one place.
- Showcase Platform: Share research findings and highlight the innovative ways student data is used to enhance learning and improve materials.
- Networking Opportunities: Make research more accessible and visible, fostering collaboration.
DREAM 2023 was not just informative but also incredibly engaging, achieving its goal of fostering collaboration and sparking new research ideas. The rich and thoughtful jam board from the event is a testament to the creative brainstorming that took place.
Our lightning talk presenters were the stars of the show. They shared fascinating insights into their research, all united by the common goal of leveraging student data to:
- Enhance students’ learning experiences and their attitudes towards statistics and data science, often perceived as challenging fields.
- Support instructors in teaching complex concepts.
- Continuously improve learning materials.
In conclusion, DREAM 2023 was a fantastic kickoff, full of networking, exploration, and knowledge sharing. We’re already dreaming up big ideas for DREAM 2024, eager to continue this exciting journey. The future is bright, and we can't wait to see where it takes us!
Details About DREAM Edu 2023
DREAM Edu 2023 brings together researchers in the CourseKata community to connect informally & exchange evidence-based insights on Thursday, November 2nd, 2023 from 9AM-11AM PT. Our program will build in ample time for casual small-group discussions, interleaved with spotlight talks from several research teams. All are welcome to join us for this FREE online workshop by registering at this link.
If you have any questions about this event, feel free to reach out to us at research@coursekata.org.
The DREAM Team
|
|
|
|
|
|
Asia Mullins, Cal State LA |
Claudia Sutter, UCLA |
Ji Son, Cal State LA |
Judith Fan, Stanford |
Program
- Warmup: Introductions + What brings you here today?
- Lightning talks
- Lightning talk #1: Icy Zhang et al.
- Lightning talk #2: Hannah Lloyd, Erik Brockbank et al.
- Lightning talk #3: Matt Jackson et al.
- Lightning talk #4: Alice Xu et al.
- Closing: What future research directions are you most excited about?
DREAM 2023 Lightning Talks
Representational-Mapping Strategies Improve Learning From an Online Statistics Textbook (slides)
Icy (Yunyi) Zhang1, Maureen E. Gray1, Alicia (Xiaoxuan) Cheng1, Ji Y. Son2, James W. Stigler1
1University of California, Los Angeles
2California State University, Los Angeles
Abstract: Using multiple representations is an important part of learning and problem-solving in science, technology, engineering and mathematics fields. At the same time, the mapping between multiple representations does not happen naturally for students. The present study developed a representation-mapping intervention designed to help students interpret, coordinate, and eventually translate across multiple representations. We integrated the intervention into an online textbook being used in a college course (i.e., CourseKata), allowing us to study its impact in a real course over an extended period of time. The findings of this study support the efficacy of the representation-mapping intervention for facilitating learning and shed light on how to implement and refine such interventions in authentic learning contexts.
Measuring Links Between Student Attitudes, Engagement, and Learning in Introductory Data Science Courses (slides)
Hannah Lloyd2, Erik Brockbank1, Zoe Tait2, Adam Bear3, Soohyun Nam Liao2, and Judith E. Fan1
1Stanford University
2University of California San Diego
3Harvard University
Abstract: Introductory data science courses provide pathways for students from diverse backgrounds to acquire computational and statistical skills. However, factors that predict achievement in these courses are not well understood. We analyzed data from 3,642 students enrolled in CourseKata courses to assess how students’ attitudes and motivation predicted their engagement and comprehension. Students who were apprehensive about the material performed less well than their peers, and this performance gap widened throughout the course. Results suggest that moderating the consequences of apprehension towards math and computer programming may be important for promoting learning outcomes for students underrepresented in math and computer science.
A “Better Book” Approach to Addressing Equity in Statistics Education
Matthew Jackson1 , Claudia C. Sutter2 , Ji Son1 , Karen Givven2
1 California State University, Los Angeles
2 University of California, Los Angeles
Abstract: We report on an intervention in an interactive introductory statistics textbook, aiming to improve equity and overall course outcomes. We center the experiences of racially minoritized students, then identify, design, and evaluate potential solutions for barriers to success. Using the “better book” approach (Stigler et al., 2019), we utilized feedback from students and instructors to redesign the text where cost perceptions peaked. Pre/Post comparisons show that perceptions of cost were reduced in all students, with racially minoritized students’ cost perceptions most dramatically improved. This suggests that removing barriers identified among minoritized students may improve the learning experience for all students.
Reading Patterns Early in an Online Statistics Textbook Predict Final Course Performance (slides)
Alice Xu1, Icy (Yunyi) Zhang1, Adam B. Blake1, and James W. Stigler1
1University of California, Los Angeles
Abstract: The rise of interactive e-textbooks in higher education emphasizes active learning but hinges on students’ effective self-regulation, which is not always achieved. There is a growing demand for a tool that can help students track their learning and provide timely feedback. This study explored the potential of machine learning in predicting student performance using CourseKata data. After the first chapter, the model identified underperforming students with 70% accuracy and 45% recall, highlighting its value for early intervention. It was found to be generalizable across diverse learners and instructors. Notably, end-of-chapter quizzes, in-text questions, and reading duration were key performance indicators.
