MoFaCTS
Use adaptive practice to study and test models of learning. Lesson access depends on available content and the deployment.
Explore the lab's research tools
LKT
Build interpretable learner models in R. MoFaCTS users can explore LKT to analyze practice histories and model learner performance.
TutorScript/SPARC
Explore a new research innovation combining explicit instructional rules, learner models, and generative AI. Instructional choices are available for inspection and study.
Adaptive practice
MoFaCTS
What it does
MoFaCTS supports research on how practice should be selected, scheduled, and adapted across different kinds of learning content. Use MoFaCTS for adaptive practice when suitable lesson content and access are available, or use the open-source code for a research deployment.
A skinned MoFaCTS implementation also supports adaptive video learning through Cognivideo.
The MoFaCTS codebase is available under the GNU Affero General Public License version 3 (AGPL-3.0-only).
Why it matters
- Supports adaptive practice experiments
- Tracks learner interaction history over time
- Connects theory to implementation and reporting
Implementation contributors
Contributors: Rusty Haner, August White, Shelbi Kuhlmann (adaptive video collaborator; skinned MoFaCTS deployment via Cognivideo).
Earlier: Andrew Tackett, Craig Kelly, Nicholas Gordon, Justin Hiller, Edward Prins III, Wes Carter, James Antoine, Ian Burns, Kenneth Brown, Andrew DiMotta, Karlos Abel, Genc Doko, Lili Wu, Giancarlo Dozzi.
MoFaCTS-related publications and presentations
- Cao, M., Pavlik, P. I., Jr., & Bidelman, G. M. (2024). Enhancing lexical tone learning for second language speakers: effects of acoustic properties in Mandarin tone perception. Frontiers in Psychology, 15, 1403816.
- Barideaux, K. J., Jr., & Pavlik, P. I., Jr. (2023). Enhancing memory recall during video lectures: does the visual display format matter? Educational Psychology, 43(6), 659-678.
- Pavlik, P. I., Jr., & Eglington, L. G. (2023). Automated Search for Logistic Knowledge Tracing Models. In Proceedings of the 16th International Conference on Educational Data Mining (pp. 17-27).
- Banker, A. M., Pavlik, P. I., Jr., Olney, A., & Eglington, L. G. (2022). Online Tutoring System (MoFaCTS) for Anatomy and Physiology: Implementation and Initial Impressions. HAPS Educator, 26(2), 44-54.
- Barideaux, K. J., Jr., & Pavlik, P. I., Jr. (2021). Can concept maps attenuate auditory distraction when studying with music? Applied Cognitive Psychology, 35(6), 1547-1558.
- Pavlik, P. I., Jr., & Eglington, L. G. (2021). The Mobile Fact and Concept Textbook System (MoFaCTS) Computational Model and Scheduling System. In CEUR Workshop Proceedings (Vol. 2895, pp. 93-107).
- Pavlik, P. I., Jr., Olney, A. M., Banker, A., Eglington, L., & Yarbro, J. (2020). The Mobile Fact and Concept Textbook System (MoFaCTS). In CEUR Workshop Proceedings (Vol. 2674, pp. 35-49).
- Eglington, L. G., & Pavlik, P. I., Jr. (2019). Predictiveness of Prior Failures Is Improved by Incorporating Trial Duration. Journal of Educational Data Mining, 11(2), 1-19.
- Murphy, C. S., & Pavlik, P. I., Jr. (2018). Effects of spacing and testing on inductive learning. Journal of Articles in Support of the Null Hypothesis, 15(1), 23-40.
- Olney, A. M., Pavlik, P. I., Jr., & Maass, J. K. (2017). Improving reading comprehension with automatically generated cloze item practice. In Artificial Intelligence in Education (Lecture Notes in Computer Science, Vol. 10331, pp. 262-273).
- Pavlik, P. I., Kelly, C., & Maass, J. K. (2016). The Mobile Fact and Concept Training System (MoFaCTS). In Lecture Notes in Computer Science (Vol. 9684, pp. 247-253).
- Maass, J. K., & Pavlik, P. I. (2016). Modeling the Influence of Format and Depth during Effortful Retrieval Practice. In Proceedings of the 9th International Conference on Educational Data Mining (pp. 143-150).
- Maass, J. K., Pavlik, P. I., Jr., & Hua, H. (2015). How spacing and variable retrieval practice affect the learning of statistics concepts. In Artificial Intelligence in Education (Lecture Notes in Computer Science, Vol. 9112, pp. 247-256).
- Barideaux, K. J., Jr., Maass, J. K., & Pavlik, P. I., Jr. (2013). A comparison of concept maps and text summaries: The benefits for learning and transfer. 54th Annual Meeting of the Psychonomic Society.
Learner modeling
LKT
What it does
Logistic Knowledge Tracing (LKT) is an R package for building and comparing interpretable models of learner performance. Researchers can combine features of practice history, success, failure, and timing to test theories and predict subsequent performance. Install LKT from CRAN and explore the tutorial to use it with your own data. MoFaCTS users are invited to explore LKT for learner modeling and analysis of practice histories.
Why it matters
- Supports modeling and feature-based analysis
- Provides reusable package-level methods and documentation
- Complements the platform side of the lab's software work
Implementation contributors
Current: Philip I. Pavlik, Jr., Luke G. Eglington.
Earlier: Liang Zhang, Leigh Harrell-Williams, Michael Yudelson, Hao Cen, Kenneth R. Koedinger, and collaborators across EDM and AIED modeling projects.
LKT-related publications and presentations
- Pavlik, P. I., Jr., & Eglington, L. G. (2023). Automated Search for Logistic Knowledge Tracing Models. In Proceedings of the 16th International Conference on Educational Data Mining (pp. 17-27).
- Eglington, L. G., & Pavlik, P. I., Jr. (2023). How to optimize student learning using student models that adapt rapidly to individual differences. International Journal of Artificial Intelligence in Education, 33(3), 497-518.
- Scruggs, R., Baker, R. S., Pavlik, P. I., Jr., McLaren, B. M., & Liu, Z. (2023). How well do contemporary knowledge tracing algorithms predict the knowledge carried out of a digital learning game? Educational Technology Research and Development, 71(3), 901-918.
- Pavlik, P. I., Eglington, L. G., & Harrell-Williams, L. M. (2021). Logistic knowledge tracing: A constrained framework for learner modeling. IEEE Transactions on Learning Technologies, 14(5), 624-639.
- Pavlik, P. I., Jr., Cen, H., & Koedinger, K. R. (2009). Performance factors analysis: A new alternative to knowledge tracing. In Proceedings of the 14th International Conference on Artificial Intelligence in Education (pp. 531-538).
- Pavlik, P., Cen, H., Wu, L., & Koedinger, K. (2008). Using item-type performance covariance to improve the skill model of an existing tutor. In Educational Data Mining 2008: 1st International Conference on Educational Data Mining, Proceedings (pp. 77-86).
New research innovation
TutorScript and SPARC
What it does
SPARC (Semantic Pages with Adaptive Rules and Cognition) is an emerging tutoring system within MoFaCTS. TutorScript is its language for expressing explicit instructional rules; the two names refer to the system and its rule language, respectively.
Learner models estimate knowledge from response evidence. TutorScript rules specify which instructional actions are eligible and how an action is selected. Generative AI then realizes the selected action as an explanation, worked example, teaching figure, or question. Separating instructional choice from language generation makes the choices available for inspection and experimental study.
Research and availability
Current research implementations support both rule-guided agent selection and student choice among instructional moves. Local browser acceptance checks have exercised these paths and saved decision histories.
These observations establish research capabilities in controlled deployments. The newer dynamic instruction policies remain restricted to administrator research use; they are not a general public feature. The wiki describes the current scope and technical details.