Cognitive Modeling


Human–Computer Interaction


Modeling human thinking on specific tasks (cognitive competencies)

Principal Metaphors

  • Knowledge is … range of established competencies
  • Knowing is … competent acting
  • Learner is … a simulated actor (computer-based simulation)
  • Learning is … developing a competency
  • Teaching is … N/A




Cognitive Modeling is concerned with simulating human thinking and problem solving, with views toward predicting human conceptual performance on focused tasks and improving human–computer interaction. Cognitive Modeling employs many different strategies, including decision-tree-like diagrams, sets of equations to simulate nonlinear systems, trainable neural nets, and interactive computer programs. Complex Systems Research serves as a core grounding to most Cognitive Modeling. Prominent subdiscourses include:
  • Bayesian Model (Bayesian Cognitive Science) (David Marr, 1970s) – an approach to modeling brain activity that is oriented by the premise that the brain is constantly operating in a space of uncertainty as it makes predictions and decisions. As a proxy for the brain’s actual coping mechanism, the model relies on a mathematical formula (developed by Thomas Bayes in the 1700s) for calculating probabilities of events as new information becomes available.
  • Diffusion Model (R. Ratcliff, 1970s) – a mathematical model for analyzing and simulating cognitive processes involved in binary (this-or-that) decisions
  • GOMS (Goals, Operators, Methods, Selection Rules) (Stuart Card, Thomas P. Moran, Allen Newell) – a theory of cognitive activity, expressed in terms of processes used to search problem spaces. Variations of GOMS include:
    • CPM-GOMS (Cognitive Perceptual Motor GOMS and Critical Path Method GOMS) (Bonnie John) – a version of GOMS that assumes multitasking behavior (rather than serial monotasking)
    • KLM-GOMS (Keystroke-Level Model) (Stuart Card, Thomas P. Moran, Allen Newell) – a model that calculates and predicts the time needed for a user to complete a familiar task
    • NGOMSL (Natural GOMS Language) (David Kieras) – a version of GOMS that offers both indications of execution times and estimations of time required to learn a system
  • Human Processor Model (Stuart Card, Thomas P. Moran, Allen Newell) – a model used to calculate and predict the time needed to perform a specific task
  • Predictive Brain (K. Nave, T. Vecchi, 2020s) – a theory that the central nervous system behaves as though it is a probability machine – that is, constantly improving its inputs and algorithms in an endless effort to reduce discrepancies between what is expected and what is actually experienced


At the moment, Cognitive Modeling is used to refer to quite a range of strategies and foci, none of which has approached the complexity of human thinking. As might be expected, however, rapid progress is being made.

Authors and/or Prominent Influences

John Anderson; Kurt VanLehn

Status as a Theory of Learning

Cognitive Modeling might be interpreted as a grounded or experimental approach to developing theories of learning.

Status as a Theory of Teaching

Cognitive Modeling is not a theory of teaching.

Status as a Scientific Theory

As noted above, there are many different strategies and foci associated with Cognitive Modeling. Most appear to maintain a strong commitment to the standards of empirical inquiry.


  • Bayesian Model (Bayesian Cognitive Science)
  • CPM-GOMS (Cognitive Perceptual Motor GOMS and Critical Path Method GOMS)
  • Diffusion Model
  • GOMS (Goals, Operators, Methods, Selection Rules)
  • Human Processor Model 
  • KLM-GOMS (Keystroke-Level Model)
  • NGOMSL (Natural GOMS Language)
  • Predictive Brain

Map Location

Please cite this article as:
Davis, B., & Francis, K. (2022). “Cognitive Modeling” in Discourses on Learning in Education.

⇦ Back to Map
⇦ Back to List