Ekaterina Shved
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Ekaterina Shved has a total of 11 co-authors. Here are the top co-authors by number of shared publications:
- Engin Bumbacher
Haute école pédagogique (HEP) Vaud, Lausanne, Switzerland
4 shared publications - Tanja Käser
EPFL
4 shared publications - Adish Singla
Max Planck Institute for Software Systems
2 shared publications - Bahar Radmehr
EPFL Switzerland
2 shared publications - Fatma Betül Güreş
ETH Zurich, Zurich, Switzerland and EPFL, Lausanne, Switzerland
2 shared publications - Paola Mejia-Domenzain
EPFL
2 shared publications
Latest Publications
- 27 June 2026
- Published by Springer Nature
- p. 154-168
Teaching Multivariational Reasoning Through AI-Guided Inquiry in Interactive Simulations
- 27 June 2026
- Published by Springer Nature
- p. 154-168
Teaching Multivariational Reasoning Through AI-Guided Inquiry in Interactive Simulations
- Open Access
- 1 April 2026
- Published by Center for Open Science in EdArXiv
Teaching Multivariational Reasoning through AI-Guided Inquiry in Interactive Simulations
Understanding how students learn and enact multivariate quantitative inquiry strategies during science inquiry remains challenging, particularly in scalable instructional settings. Although strategies such as the extended Control of Variables Strategy (CVS++) and the General Principle Strategy (GPS) are known to support multivariate reasoning, less is known about how instruction can foster their uptake and sustained use during inquiry.We investigate this issue using a rule-based AI chatbot that scaffolds strategy-specific reasoning during simulation-based inquiry, providing context-sensitive prompts to guide students in enacting either CVS++ or GPS. In a between-groups experiment, 104 chemistry apprentices engage in a three-phase inquiry sequence: inquiry with AI-supported guidance Learning Phase, independent inquiry in the same simulation with a new variable Near Transfer, and inquiry in a novel simulation Far Transfer. Students’ inquiry actions and strategy enactment were captured through fine-grained interaction log data. Results show that CVS++ leads to higher post-instruction performance than GPS, driven primarily by single-variable reasoning items, while no group differences emerge in Near or Far Transfer. Behavioral analyses indicate that students largely enact the instructed strategies in phases close to instruction, but explicit strategy use declines in the novel simulation, particularly for GPS. Across phases, higher learning is consistently associated with more systematic inquiry behaviors, including coordinated exploration, recording, and analysis actions, rather than strategy use alone.
- 1 April 2026
- Published by Center for Open Science in EdArXiv
Open AccessTeaching Multivariational Reasoning through AI-Guided Inquiry in Interactive Simulations
Understanding how students learn and enact multivariate quantitative inquiry strategies during science inquiry remains challenging, particularly in scalable instructional settings. Although strategies such as the extended Control of Variables Strategy (CVS++) and the General Principle Strategy (GPS) are known to support multivariate reasoning, less is known about how instruction can foster their uptake and sustained use during inquiry.We investigate this issue using a rule-based AI chatbot that scaffolds strategy-specific reasoning during simulation-based inquiry, providing context-sensitive prompts to guide students in enacting either CVS++ or GPS. In a between-groups experiment, 104 chemistry apprentices engage in a three-phase inquiry sequence: inquiry with AI-supported guidance Learning Phase, independent inquiry in the same simulation with a new variable Near Transfer, and inquiry in a novel simulation Far Transfer. Students’ inquiry actions and strategy enactment were captured through fine-grained interaction log data. Results show that CVS++ leads to higher post-instruction performance than GPS, driven primarily by single-variable reasoning items, while no group differences emerge in Near or Far Transfer. Behavioral analyses indicate that students largely enact the instructed strategies in phases close to instruction, but explicit strategy use declines in the novel simulation, particularly for GPS. Across phases, higher learning is consistently associated with more systematic inquiry behaviors, including coordinated exploration, recording, and analysis actions, rather than strategy use alone.
- 21 July 2025
- Published by Springer Nature
- p. 94-102
ClickSight: Interpreting Student Clickstreams to Reveal Insights on Learning Strategies via LLMs
- 21 July 2025
- Published by Springer Nature
- p. 94-102
ClickSight: Interpreting Student Clickstreams to Reveal Insights on Learning Strategies via LLMs
- Open Access
ClickSight: Interpreting Student Clickstreams to Reveal Insights on Learning Strategies via LLMs
Clickstream data from digital learning environments offer valuable insights into students' learning behaviors, but are challenging to interpret due to their high dimensionality and granularity. Prior approaches have relied mainly on handcrafted features, expert labeling, clustering, or supervised models, therefore often lacking generalizability and scalability. In this work, we introduce ClickSight, an in-context Large Language Model (LLM)-based pipeline that interprets student clickstreams to reveal their learning strategies. ClickSight takes raw clickstreams and a list of learning strategies as input and generates textual interpretations of students' behaviors during interaction. We evaluate four different prompting strategies and investigate the impact of self-refinement on interpretation quality. Our evaluation spans two open-ended learning environments and uses a rubric-based domain-expert evaluation. Results show that while LLMs can reasonably interpret learning strategies from clickstreams, interpretation quality varies by prompting strategy, and self-refinement offers limited improvement. ClickSight demonstrates the potential of LLMs to generate theory-driven insights from educational interaction data.
Open AccessClickSight: Interpreting Student Clickstreams to Reveal Insights on Learning Strategies via LLMs
Clickstream data from digital learning environments offer valuable insights into students' learning behaviors, but are challenging to interpret due to their high dimensionality and granularity. Prior approaches have relied mainly on handcrafted features, expert labeling, clustering, or supervised models, therefore often lacking generalizability and scalability. In this work, we introduce ClickSight, an in-context Large Language Model (LLM)-based pipeline that interprets student clickstreams to reveal their learning strategies. ClickSight takes raw clickstreams and a list of learning strategies as input and generates textual interpretations of students' behaviors during interaction. We evaluate four different prompting strategies and investigate the impact of self-refinement on interpretation quality. Our evaluation spans two open-ended learning environments and uses a rubric-based domain-expert evaluation. Results show that while LLMs can reasonably interpret learning strategies from clickstreams, interpretation quality varies by prompting strategy, and self-refinement offers limited improvement. ClickSight demonstrates the potential of LLMs to generate theory-driven insights from educational interaction data.
- 2 July 2024
- Published by Springer Nature
- p. 482-496
Teaching and Measuring Multidimensional Inquiry Skills Using Interactive Simulations
- 2 July 2024
- Published by Springer Nature
- p. 482-496
Teaching and Measuring Multidimensional Inquiry Skills Using Interactive Simulations
- Open Access
- 8 May 2024
- Published by Center for Open Science in EdArXiv
Teaching and Measuring Multidimensional Inquiry Skills using Interactive Simulations
Interactive simulations play a significant role in science education, serving as a platform for inquiry-based learning and fostering the development of scientific knowledge and skills. However, teaching and quantitatively measuring inquiry strategies has proven to be challenging due to their complex and inherently multidimensional nature. Our study goes beyond the prevalent focus on the Control of Variables Strategy (CVS) in prior work by incorporating additional relevant inquiry strategies in both teaching and measurement: exploring the variable range and conducting experiments under optimal conditions. We tested two different instructional approaches to jointly teach the three strategies by focusing either on data collection or on data interpretation. 161 chemistry apprentices were randomly assigned to one of the two instructional conditions or a control group without instruction and engaged in experimentation using an interactive simulation. In order to analyze joint strategy use, we applied a multi-step clustering method to students' log data that helped identify multidimensional student profiles of inquiry strategies. We found four profiles that related differently to conceptual learning, suggesting that combining strategies is more effective for conceptual learning than utilizing them individually. We also found that students instructed on data collection increased the use of strategy combinations with an emphasis on CVS. This suggests a potential avenue for assessing instruction efficacy, indicating that the impact may be strategy-specific. Source code and materials are released at https://github.com/epfl-ml4ed/inquiry-skills.
- 8 May 2024
- Published by Center for Open Science in EdArXiv
Open AccessTeaching and Measuring Multidimensional Inquiry Skills using Interactive Simulations
Interactive simulations play a significant role in science education, serving as a platform for inquiry-based learning and fostering the development of scientific knowledge and skills. However, teaching and quantitatively measuring inquiry strategies has proven to be challenging due to their complex and inherently multidimensional nature. Our study goes beyond the prevalent focus on the Control of Variables Strategy (CVS) in prior work by incorporating additional relevant inquiry strategies in both teaching and measurement: exploring the variable range and conducting experiments under optimal conditions. We tested two different instructional approaches to jointly teach the three strategies by focusing either on data collection or on data interpretation. 161 chemistry apprentices were randomly assigned to one of the two instructional conditions or a control group without instruction and engaged in experimentation using an interactive simulation. In order to analyze joint strategy use, we applied a multi-step clustering method to students' log data that helped identify multidimensional student profiles of inquiry strategies. We found four profiles that related differently to conceptual learning, suggesting that combining strategies is more effective for conceptual learning than utilizing them individually. We also found that students instructed on data collection increased the use of strategy combinations with an emphasis on CVS. This suggests a potential avenue for assessing instruction efficacy, indicating that the impact may be strategy-specific. Source code and materials are released at https://github.com/epfl-ml4ed/inquiry-skills.