People cannot tell you why they did it.

Nisbett and Wilson: people often cannot accurately report what influenced their judgements, and what they offer instead is a plausible theory, delivered sincerely. In their shop study the rightmost of four identical items was chosen close to four to one, nobody mentioned position, and asked directly, virtually all of them denied it. Schooler and Engstler-Schooler: people who described a face in words were worse at recognising it afterwards.

Completely, and my Insight module says it without the citation. When we ask people why they do the things they do, we are asking them to give voice to their intuition, which by definition they cannot, so they think of an answer, and that answer is different from what their intuitive response would have been.

They establish that introspection is unreliable and stop there. I still needed the answer, because a client is paying for one.

Because the practice had to build a working method instead. Observe rather than ask. Gather stories rather than reasons. Talk in verbs, and do the interpreting yourself afterwards. And on the second paper, do not ask somebody to headline an emerging idea, because summarising it swaps the idea for a poorer description of it in front of an audience. Later replication work puts that effect smaller than the original, so I rely on the direction, which has held, and I say so.

People cannot tell you why they did it. So stop asking, and go and watch instead.