Description
Compares the sequences of fixations across users and designs with string-edit or vector methods, revealing reading strategies that heatmaps conceal.
Heatmaps aggregate where people look but discard order and timing. Scanpath analysis treats fixation sequences as strings over areas of interest (AOIs) or as vectors, then computes similarity: Levenshtein edit distance, ScanMatch's Needleman-Wunsch alignment with substitution matrices and duration, or MultiMatch's five dimensions (shape, direction, length, position, duration). Clustering similar scanpaths reveals distinct strategies (e.g. F-pattern versus layer-cake reading). Transition matrices and entropy measures show how structured the viewing is. This answers questions like 'do users read the warning before the button?' and 'does the redesign create one consistent strategy?'.
Attributes
- Discipline
- Neuro, Biometrics & Affect
- Loss type
- Latent
- Stage
- Evaluation
- Data
- Quantitative
- Effort
- Weeks
Steps
- Define AOIs on stimuli, and ensure eye-tracking data quality (accuracy below about 1 degree, low data loss).
- Extract fixation sequences per participant and task, collapsing consecutive fixations within an AOI if appropriate.
- Compute pairwise similarity using ScanMatch or MultiMatch, and transition matrices with stationary and transition entropy.
- Cluster scanpaths to identify strategies; compare within-design and between-design similarity with permutation tests.
- Check whether critical sequences occur (e.g. warning AOI fixated before action AOI) and their frequency.
- Link strategies to task outcomes such as errors or time.
Sources
- Noton & Stark, Scanpaths in eye movements during pattern perception (1971), Science
- Cristino et al., ScanMatch: A novel method for comparing fixation sequences (2010), Behavior Research Methods
- Holmqvist et al., Eye Tracking: A Comprehensive Guide to Methods and Measures (2011)
Also known as
- Scanpath comparison
- ScanMatch
- MultiMatch
- String-edit scanpath analysis
Categories
- Analytics & Behaviour
- Neuro, Biometrics & Affect