Logistic knowledge tracing (LKT): a symbolic notation framework for specifying logistic regression learner models¶

Theory · All theories
Evidence · 1 claim (1 for) · 1 study (1 design), q2 · 0 of 1 report an effect size · 1 claim rests on one study

Description¶

LKT is a theoretically motivated framework for systematic specification and evaluation of student models for adaptive instructional environments, built on logistic regression. Its strength is a symbolic notation system in which models are composed of terms, each with a feature applied at a component level (student, item, KC), including intercept, linear, and nonlinear features such as recency, decay, and spacing. The framework subsumes extant models like AFM, PFA, and R-PFA as special cases and facilitates comparing their strengths and weaknesses.

Design Implications¶

Context¶

Requirements¶

• A sequence of learner event data with subject and correctness columns, minimally in DataShop format

Constraints¶

• The article intentionally limits scope to logistic regression models and does not compare against BKT or deep knowledge tracing

Target Learners¶

• students using adaptive learning technology systems

Target Learning Objectives¶

• predicting correctness of subsequent practice items to guide pedagogical decisions

Claims¶

• No Single Learner Model Best Across Datasets [+M]

Related Theories¶

Examples¶

-

Key Sources¶

• Philip I. Pavlik, Jr., Luke G. Eglington, and Leigh M. Harrell-Williams. (2021). Logistic Knowledge Tracing: A Constrained Framework for Learner Modeling. IEEE Transactions on Learning Technologies. https://doi.org/10.1109/TLT.2021.3128569