Three short math/word problems with answers that "feel" right but aren't. The original Frederick (2005) CRT plus three bonus questions. Type your first instinct — then your final answer after reflection.
What makes this one different
Standard CRT just counts correct answers. We capture your first instinct separately from your final answer. The gap between them measures how well you can override an intuitive response — that's the actual cognitive signal.
Reflection gain = correct final − correct first instinct. If it's positive, you actually overrode impulses. If zero, your instincts are either already good or you didn't stop to think.
No sessions yet.
Shane Frederick's 2005 CRT is famous for a reason: it isolates cognitive reflection — the ability to recognize that an intuitive answer is wrong and override it. It's just three questions, but it correlates with a wide range of rational-thinking measures: probabilistic reasoning, expected-value calculations, even some religious-belief statistics.
3/3 on the original Frederick CRT is the median for MIT students. 0/3 is normal for the general population.
Reflection gain > 0 means you actually overrode impulses. = 0 means your instincts are well-calibrated (or you didn't stop to think).
That's the point. They expose the gap between fast intuitive thinking and slower reflective thinking. If a question doesn't trick you, your reflective system is online.
Either you've seen them before, or your intuitive system happens to match the correct mathematical answer — both fine. The reflection gain metric still applies to questions where your first instinct was wrong.