Description
Compares how cues relate to a true criterion with how people actually weight those cues, quantifying judgement accuracy, consistency and cue misuse.
The lens model places a distal criterion (e.g. actual trustworthiness of a seller) on one side and a person's judgement on the other, connected through a set of proximal cues (ratings, photos, response time). Regressing the criterion on cues gives ecological validities; regressing judgements on cues gives cue utilisation. The lens model equation decomposes achievement (the correlation between judgement and criterion) into environmental predictability, judge consistency and matching of cue weights. For UX, this precisely diagnoses whether users misjudge because the interface presents weak cues, because they weight cues wrongly, or because they are inconsistent, and so tells you whether to change the cues or help users use them.
Attributes
- Discipline
- Cognitive Psychology & Perception
- Loss type
- Unmigrated
- Stage
- Definition, Evaluation
- Data
- Quantitative
- Effort
- Weeks
Steps
- Define the criterion users are judging and obtain ground-truth values for a set of cases (e.g. 50-100 profiles, products or messages).
- Identify the cues available in the interface and code their values for each case.
- Have participants judge each case; include repeated cases to estimate consistency.
- Regress criterion on cues (environment side) and each participant's judgements on cues (judge side).
- Compute the lens model equation components: achievement, environmental predictability, judge consistency (Rs) and knowledge (G).
- Redesign: add or strengthen valid cues, de-emphasise invalid but salient ones, and retest achievement.
Sources
- Brunswik, Perception and the Representative Design of Psychological Experiments (1956)
- Hammond & Stewart (eds.), The Essential Brunswik (2001)
- Cooksey, Judgment Analysis: Theory, Methods, and Applications (1996)
Also known as
- Lens model
- Social judgement theory
- Probabilistic functionalism
- Judgement analysis
Categories
- User Research
- Cognitive Psychology & Perception