Methodology
How Wouch reads a pattern and what it refuses to claim.
We can\'t show you the engine the way we weight and combine signals is the part that\'s genuinely ours. But we can be completely clear about the method: what it\'s built on, what it does, and the lines it will not cross. Restraint is the rigour.
The shape of the method
We hold beliefs as probabilities, and let your answers move them.
Most assessments give you a fixed result: a type, a score, a box. Wouch doesn't work that way. We use Bayesian inference which is a precise way of saying we start with an honest amount of uncertainty about you, and update it gently as you tell us more. Nothing is decided from a single answer. A picture forms across many, and it keeps the right amount of doubt attached to it.
What we build on
Established, peer-reviewed measures not a made-up quiz.
Attachment — adapted from the Experiences in Close Relationships–Revised (ECR-R; Fraley, Waller & Brennan, 2000), measuring attachment-related anxiety and avoidance. Used with permission.
Loneliness — items from the UCLA Loneliness Scale, Version 3 (Russell, 1996) and its brief 3-item form (Hughes, Waite, Hawkley & Cacioppo, 2004). Used with permission.
Distress screening — the PHQ-4 (Kroenke, Spitzer, Williams & Löwe, 2009), a brief screen for low mood and anxiety. Public domain; no permission required.
How these measures are combined, weighted and updated over time is proprietary to Wouch. The measures are open and cited; the engine is ours
We would rather show you complexity you can navigate than a number you\'d only perform for.Why there is no score