More Like this

1. Scalable and Equitable Math Problem Solving Strategy Prediction in Big Educational Data

   Shakya, A.; Rus, V.; Venugopal, D. (July 2023, Proceedings of the 16th International Conference on Educational Data Mining, International Educational Data Mining Society)

   Understanding a student's problem-solving strategy can have a significant impact on effective math learning using Intelligent Tutoring Systems (ITSs) and Adaptive Instructional Systems (AISs). For instance, the ITS/AIS can better personalize itself to correct specific misconceptions that are indicated by incorrect strategies, specific problems can be designed to improve strategies and frustration can be minimized by adapting to a student's natural way of thinking rather than trying to fit a standard strategy for all. While it may be possible for human experts to identify strategies manually in classroom settings with sufficient student interaction, it is not possible to scale this up to big data. Therefore, we leverage advances in Machine Learning and AI methods to perform scalable strategy prediction that is also fair to students at all skill levels. Specifically, we develop an embedding called MVec where we learn a representation based on the mastery of students. We then cluster these embeddings with a non-parametric clustering method where each cluster contains instances that have approximately symmetrical strategies. The strategy prediction model is trained on instances sampled from these clusters ensuring that we train the model over diverse strategies. Using real world large-scale student interaction datasets from MATHia, we show that our approach can scale up to achieve high accuracy by training on a small sample of a large dataset and also has predictive equality, i.e., it can predict strategies equally well for learners at diverse skill levels. 

   more » « less
2. Sensor-Free Predictive Models of Affect in an Online Learning Environment

   Harrison, A (January 2018, Proceedings of the Eleventh International Conference on Educational Data Mining)

   A significant amount of research has illustrated the impact of student emotional and affective state on learning outcomes. Just as human teachers and tutors often adapt instruction to accommodate changes in student affect, the ability for computer-based systems to similarly become affect-aware, detecting and personalizing instruction in response to student affective state, could significantly improve student learning. Personalized and affective interventions in tutoring systems can be realized through affect-aware learning technologies to deter students from practicing poor learning behaviors in response to negative affective states and to optimize the amount of learning that occurs over time. In this paper, we build off previous work in affect detection within intelligent tutoring systems (ITS) by applying two methodologies to develop sensor-free models of student affect with only data recorded from middle-school students interacting with an ITS. We develop models of four affective states to evaluate and determine significant predictors of affect. Namely, we develop a model which discerns students’ reported interest significantly better than majority class. 

   more » « less
3. Eliciting Proactive and Reactive Control During Use of an Interactive Learning Environment

   Unal, D. S.; Arrington, C. M.; Solovey, E.; Walker, E. (January 2023, Artificial Intelligence in Education (AIED 2023))

   Mendez, G.; Matsuda, N.; Santos, O. C.; Dimitrova, V. (Ed.)

   The dual mechanisms of control framework describes two modes of goal-directed behavior: proactive control (goal maintenance) and reactive control (goal activation on task demands). Although these mechanisms are relevant to learner behaviors during interaction with intelligent tutoring systems (ITS), their relation to ITSs is under-researched. We propose a manipulation to induce proactive or reactive control during interaction with an online tutoring system. We present two experiments where students solved problems using either proactive or reactive control. Study 1 validates the manipulation by investigating behavioral measures that reflect usage of the intended strategy and assesses whether either mode impacted learning. Study 2 investigates if alternating between control modes during problem solving affects student performance. 

   more » « less
4. Exploring Simultaneous Knowledge and Behavior Tracing

   https://doi.org/10.5281/zenodo.12730000  

   Zhao, Siqian; Sahebi, Sherry (July 2024, International Educational Data Mining Society)

   Benjamin, Paaßen; Carrie, Demmans Epp (Ed.)

   Knowledge Tracing (KT) focuses on quantifying student knowledge according to the student's past performance. While KT models focus on modeling student knowledge, they miss the behavioral aspect of learning, such as the types of learning materials that the students choose to learn from. This is mainly because traditional knowledge tracing (KT) models only consider assessed activities, like solving questions. Recently, there has been a growing interest in multi-type KT which considers both assessed and non-assessed activities (like video lectures). Since multi-type KT models include different learning material types, they present a new opportunity to investigate student behavior, as in the choice of the learning material type, along with student knowledge. We argue that student knowledge can affect their behavior, and student interest in learning materials may affect their knowledge. In this paper, we model the relationship between students' knowledge states and their choice of learning activities. To this end, we propose Pareto-TAMKOT which frames the simultaneous learning of student knowledge and behavior as a multi-task learning problem. It employs a transition-aware multi-activity KT method for two objectives: modeling student knowledge and student behavior. Pareto-TAMKOT uses the Pareto Multi-task learning algorithm (Pareto MTL) to solve this multi-objective optimization problem. We evaluate Pareto-TAMKOT on one real-world dataset, demonstrating the benefit of approaching student knowledge and behavior modeling as a multi-task learning problem. 

   more » « less
5. Aligning tutor discourse supporting rigorous thinking with tutee content mastery for predicting math achievement

   Abdelshiheed, M; Jacobs, JK; D’Mello, S (July 2024, Proceedings of the International Conference on Artificial Intelligence in Education)

   Olney, AM; Chounta, IA; Liu, Z; Santos, OC; Bittencourt, II (Ed.)

   This work investigates how tutoring discourse interacts with students’ proximal knowledge to explain and predict students’ learning outcomes. Our work is conducted in the context of high-dosage human tutoring where 9th-grade students attended small group tutorials and individually practiced problems on an Intelligent Tutoring System (ITS). We analyzed whether tutors’ talk moves and students’ performance on the ITS predicted scores on math learning assessments. We trained Random Forest Classifiers (RFCs) to distinguish high and low assessment scores based on tutor talk moves, student’s ITS performance metrics, and their combination. A decision tree was extracted from each RFC to yield an interpretable model. We found AUCs of 0.63 for talk moves, 0.66 for ITS, and 0.77 for their combination, suggesting interactivity among the two feature sources. Specifically, the best decision tree emerged from combining the tutor talk moves that encouraged rigorous thinking and students’ ITS mastery. In essence, tutor talk that encouraged mathematical reasoning predicted achievement for students who demonstrated high mastery on the ITS, whereas tutors’ revoicing of students’ mathematical ideas and contributions was predictive for students with low ITS mastery. Implications for practice are discussed. 

   more » « less