Systematic Error Tracing (SET) model¶

Theory · All theories
Evidence · no claims cited

Description¶

The SET model is a knowledge-based recommendation system that traces the latent causes of a student's systematic errors using a bipartite graph linking causes to errors, derived from the Ising model. It computes cause probabilities from the relative number of links between a cause and the observed errors, which the authors note is "in accordance with Luce's choice axiom," and can be generalized to a given number of co-occurring causes. It is deliberately not tied to a specific theoretical framework of error origins and is designed for easy implementation in online learning systems, updating after each newly observed error response.

Design Implications¶

Context¶

Requirements¶

• A known error-to-cause mapping captured in a bigraph, with clearly identifiable errors and mappable possible causes

Constraints¶

• The SET model cannot predict a cause that is not part of the bigraph
• It is confined to items that elicit an erroneous response
• It exhibits a bias to appeal to majority, which can misidentify slips as dominant causes

Target Learners¶

• primary school children practicing arithmetic

Target Learning Objectives¶

• diagnosis of systematic error causes in single-digit multiplication to guide intervention recommendations

Claims¶

• Set Outperforms Majority Vote Beyond Top Recommendation [+M]

Related Theories¶

Examples¶

• Use traced error causes as personalized intervention recommendations in online learning systems

Key Sources¶

• Savi, A. O., Deonovic, B. E., Bolsinova, M., van der Maas, H. L. J., & Maris, G. K. J. (2021). Tracing Systematic Errors to Personalize Recommendations in Single Digit Multiplication and Beyond. Journal of Educational Data Mining, Volume 13, No 4. https://jedm.educationaldatamining.org/index.php/JEDM/article/view/382