Magnitude Estimation and Stevens' Power Law

Description

Participants assign numbers proportional to perceived magnitude, yielding ratio-scale data and power-law exponents that describe how perception scales with the stimulus.

In magnitude estimation, participants see a reference stimulus (modulus) and assign numbers to others in proportion: twice as intense gets twice the number. Geometric means across participants, plotted against stimulus intensity on log-log axes, typically give a straight line whose slope is the power-law exponent (psi = k x phi^n). Exponents differ by dimension: perceived brightness is compressive (about 0.33), electric shock expansive (about 3.5). Magnitude estimation was used in early HCI and usability research (McGee, 2003, applied it to usability judgements), offering ratio-level measurement of subjective experience that Likert scales cannot provide.

Attributes

Loss type
Extinct
Stage
Evaluation
Effort
Days

Steps

  1. Select stimuli spanning the range of interest (e.g. 6-10 levels of load time or visual clutter).
  2. Present a modulus stimulus with an assigned or free number, then present test stimuli in random order with repeats.
  3. Instruct participants to assign numbers proportional to perceived magnitude, allowing any positive number.
  4. Normalise each participant's responses and compute geometric means per stimulus.
  5. Fit a power function by linear regression on log-log values and report the exponent with confidence intervals.
  6. Use the exponent to predict perceived improvement from physical changes, or compare designs on ratio-scale judgements.

Sources

  • Stevens, Psychophysics: Introduction to Its Perceptual, Neural, and Social Prospects (1975)
  • Stevens, On the psychophysical law (1957), Psychological Review
  • McGee, Usability magnitude estimation (2003), Proceedings of HFES
  • Quantifying the User Experience (Sauro & Lewis)

Also known as

  • Magnitude estimation
  • Ratio scaling
  • Stevens' power law
  • Cross-modality matching

Categories

  • User Research
  • Cognitive Psychology & Perception