GOMS (Goals, Operators, Methods, Selection rules) Models: Analyzing Human-Computer Interaction and Predicting User Performance

Explore the GOMS family of models for analyzing user behavior and predicting task completion times in human-computer interaction (HCI). This guide explains the core GOMS components, details specific models like KLM and CMN-GOMS, and demonstrates their application in evaluating user interface designs and usability.

GOMS Family of Models: Analyzing Human-Computer Interaction

Introduction to GOMS Models

The GOMS (Goals, Operators, Methods, and Selection rules) family of models is a framework in cognitive psychology and human-computer interaction (HCI) for modeling the cognitive processes involved in performing tasks using a computer system. Developed by Card, Moran, and Newell in the 1980s, GOMS provides a way to represent how a user plans and executes actions to achieve a goal. Several models extend the original GOMS framework, including the Keystroke Level Model (KLM) and CMN-GOMS.

Key Components of GOMS Models

GOMS models use four key components to represent user behavior:

  • Goals: The user's objectives (high-level goals are broken down into sub-goals).
  • Operators: The basic physical and cognitive actions a user performs (e.g., keystrokes, mouse movements, mental operations). These are considered atomic—indivisible units.
  • Methods: Sequences of operators used to achieve a goal or subgoal.
  • Selection Rules: Rules determining which method to use when multiple methods are available for a goal.

1. Keystroke Level Model (KLM)

KLM is a simplified GOMS model focusing on predicting task completion times. It breaks down tasks into basic actions (keystrokes, mouse movements, etc.) and assigns time values to each action. This allows for a quantitative estimation of task performance.

KLM Operators and Time Estimates:

  • K (Keystroke): Time to press a key (0.2 seconds).
  • P (Pointing): Time to move the mouse pointer (1.1 seconds).
  • H (Home-hand): Time to move hand from mouse to keyboard (0.4 seconds).
  • M (Mental): Mental preparation time (variable, depends on task complexity).
  • B (Button Press): Time to press a mouse button (0.1 seconds).
  • R (System Response): System response time (often negligible in modern systems).

(Further details on the KLM model and its application in predicting task completion times, including an example, would be added here.)

2. CMN-GOMS

CMN-GOMS (Card, Moran, Newell GOMS) is a more comprehensive GOMS model. It uses the same four constructs (goals, operators, methods, selection rules) but provides a more detailed and flexible way to model complex user behavior. The operators in CMN-GOMS are more broadly defined than in KLM.

CMN-GOMS Constructs:

  • Goals: Hierarchical representation of user objectives.
  • Operators: Basic actions (broader definition than in KLM).
  • Methods: Sequences of operators for achieving subgoals.
  • Selection Rules: Determine which method to use when multiple options are available.

CMN-GOMS: A Hierarchical Model of Human-Computer Interaction

Introduction to CMN-GOMS

CMN-GOMS (Card, Moran, Newell GOMS) is a widely used model in cognitive psychology and human-computer interaction (HCI) for analyzing user behavior. It provides a more detailed and flexible approach than the Keystroke Level Model (KLM), another model in the GOMS family. CMN-GOMS helps researchers and designers understand how users plan and execute tasks, improving the design and usability of user interfaces. The model uses four key constructs: Goals, Operators, Methods, and Selection rules, to represent user behavior.

CMN-GOMS vs. KLM: Key Differences

While CMN-GOMS and KLM share similarities in their use of the four GOMS constructs, they differ in several crucial ways:

  • Operators: KLM has a predefined set of seven operators, while CMN-GOMS allows for the definition of operators at any level of cognitive detail. This flexibility enables more nuanced modeling.
  • Cognitive Modeling: KLM assumes a linear sequence of actions, while CMN-GOMS allows for a hierarchical representation of cognitive processes. This enables the representation of more complex task structures.

CMN-GOMS Constructs: A Deeper Dive

1. Goals

Goals represent the user's high-level objectives. The hierarchical nature of CMN-GOMS allows for representing complex tasks by breaking down high-level goals into sub-goals.

2. Operators

Operators are the basic actions the user performs to work toward a goal. These can be physical actions (like keystrokes) or cognitive actions (like making a decision).

3. Methods

Methods describe the specific sequence of operators used to achieve a subgoal. Methods organize operators into a structured plan for accomplishing sub-tasks within a larger task.

4. Selection Rules

Selection rules determine which method to use when multiple methods are available for a given subgoal. They provide a way to model the user's decision-making process when choosing between alternative actions.