A collection of theories in HCI

This page gives a one-sentence pointer to theoretical models in HCI, and points to people at our chair who have some experience using those theories in their research.

Disclaimer: The following descriptions are often copied directly or in slightly modified form from the linked source, because they are well-described; i.e., you may find extensive plagiarism throughout this page. We don't claim any originality here. :D


Low-level theories

Concerning individual perception and behavior.

Motor system

  • Fitts's law predicts movement time in target acquisition based on distance to and width of the target.
    (dis1, Chat, Kashyap)
  • Steering law predicts movement time required to steer through a 2-dimensional tunnel based on the (possibly dynamic) width of the path.
    (Rene, Simon)
  • Weber-Fechner law indicates that the just-noticeable difference between two stimuli is log-proportional to the magnitude of the stimuli.
    (Christian)
  • GOMS (Goals, Operators, Methods, and Selection rules) predicts expert execution and learning times for tasks based on a hierarchical decomposition into basic actions.
    (DIS1)


Cognitive system

  • Hick's law predicts time to make a decision based on the number of choices present.
    ( )
  • Decay theory indicates that the more information chunks need to be remembered in short-term memory, the less likely they can be recalled as time passes.
    ( )
  • Information Foraging describes how people search for information. We have a good book on that.
    ( )
  • Media Equation is a general communication theory that shows that people tend to treat computers and other new media as if they were real people or places.
    ( )
  • Gestalt theory explains what cues lead humans to perceive a set of objects as belonging together.
    (DIS1)



High level theories


Concerning behavior of multiple people.

  • Activity theory explains how context, social artifacts, and social organization shape user actions.
    ( )
  • Distributed Cognition by Ed Hutchins states that cognition is better explained as a distributed phenomenon involving multiple people and artefacts.
    (Carsten; we also have good connections to the inventor's lab at UCSD.)


Concerning human in psychology

  • Ecological approach by Gibson stresses the relation between the organism and its environment and the relation between perception and action.
    ( )



Books

  • The Psychology of Human-Computer Interaction (Card, Moran, Newell, 1983) - at our chair. First established using psychological models for HCI, includes details on GOMS and other theories.
  • HCI Models, Theories, and Frameworks (John M. Carroll, 2003) is a more recent collection of HCI theories; unfortunately it's out of print but you can see some content online.



Constructing mathematical models

In his CHI ’14 note, Oulasvirta provided a program that searches the space of possible nonlinear combination for these parameters and allow generating the model with best fit. The smart thing about this is that he narrowed the search space down by using characteristics of HCI models (few independent variables, starting with linear operations, favoring some nonlinear operations, created from averaged point data).
  1. Do your experiment to get the data of the behavior that you want to model.
  2. Read about basics of modeling from reference 1,9
  3. Read Oulasvirta's note (4 pages): http://dx.doi.org/10.1145/2556288.2556999
  4. Use Oulasvirta's code to search for models. (http://www.mpi-inf.mpg.de/~oantti/nonlinearmodeling/)
  5. Beautify the resulting model and provide explanations for the terms and parameters.
Details