We don't ask how you think — we watch you thinking. And that changes everything.
Metacognition is poorly measured by questionnaires, because the person who doesn't enjoy thinking also won't spend much time reflecting on whether they enjoy thinking. The alternative: a guided AI conversation that observes your cognitive style in action.
Three established psychological constructs (Need for Cognition, ambiguity tolerance, attributional style) plus a map of communication aesthetic preferences. All extracted from natural conversation — no checkboxes.
If you give someone an ambiguous dilemma and observe their response, you're measuring ambiguity tolerance ecologically. Whether they seek more information, tolerate uncertainty, propose conditional strategies, or demand a binary answer — each response is a data point.
The pleasure of thinking. Do you enjoy chewing on complex problems or prefer the direct answer? Determines whether AI explains the "why" or tells you the "what."
Can you hold multiple hypotheses simultaneously or do you need to close on an answer? Determines whether AI says "it depends, and here are the nuances" or "the answer is X."
Metaphors vs. data, narrative vs. analysis, prose vs. bullets, humor vs. seriousness. The preferences that make AI sound like a chosen interlocutor, not a chatbot.
The Need for Cognition (Cacioppo & Petty, 1982) is typically measured with the 18-item NCS, with alphas above .85 and stable unifactorial structure. Ambiguity tolerance (Budner, 1962) is a robust predictor of information processing preferences.
Afini's innovation: instead of measuring these constructs with questionnaires, it extracts them from real conversational behavior. The conversation is processed with an extraction prompt that identifies cognitive and aesthetic patterns. It's the layer with highest ecological validity — and what most differentiates Afini from any competitor.
Key references
Cacioppo, J. T. & Petty, R. E. (1982). The need for cognition. JPSP, 42, 116-131. · Budner, S. (1962). Intolerance of ambiguity as a personality variable. Journal of Personality, 30, 29-50.
Cognitive style is the ultimate fine-tuning. If Big Five sets the tone and attachment the temperature, cognitive style calibrates the texture: metaphors or data? Flowing prose or scannable bullets? Nuances or conclusions?
"Climate change is a complex phenomenon involving multiple factors: greenhouse gas emissions, deforestation, land use changes, and climate system feedbacks..."
"Think of it as an engineering problem with multiple feedback loops — which I know is how you like to think. CO₂ is the obvious input, but the interesting loops are the ones most people ignore: ice albedo, permafrost methane, ocean absorption reaching saturation. There's no magic lever, and I know that doesn't frustrate you — it fascinates you."
Same information. But one is packaged for a generic reader and the other for someone with high NFC who thinks in systems and tolerates complexity.
No questionnaires. 15 minutes of conversation. And AI will learn whether you prefer metaphors or data, prose or bullets, nuance or direct answers.