All best practices
Behavioural Science
May 31, 2026 · 3 min read
Most decisions are made before the conscious mind gets involved.

Measure the Automatic Response: The Decisions People Can’t Explain

A great deal of choice is fast, automatic, and emotional — driven by associations the respondent can’t narrate. If your method only captures what people can put into words, you are measuring the smaller half.
The practice

Measure the automatic response, not only the articulated one. Reaction time and instinctive choice reveal the fast, associative drivers a questionnaire can’t reach.


Fast and slow — a metaphor, not a map

It is useful to picture two modes of thinking. One is fast, automatic, and effortless — the snap judgement, the gut pull toward a pack you have seen a thousand times. The other is slow, deliberate, and effortful — the reasoning you do when you sit down to justify a choice. These are convenient labels for styles of processing, not two literal systems wired into the head¹ — but, used carefully, they name a real and repeatedly observed difference. Most everyday decisions run on the fast, automatic style, and only get handed to the deliberate one when something forces the issue.

The part a questionnaire misses

A survey speaks almost entirely to slow, deliberate processing. It asks the respondent to stop, consider, and explain — which is exactly the mode the real decision did not use. Whatever drove the fast, automatic response is largely invisible to self-report, because the respondent has no reliable access to it.² They are not withholding it; they genuinely cannot see it to tell you.

Reading the automatic response in practice

The way in is to measure the response itself rather than the report of it. Reaction time, forced fast choice, and implicit association tasks all capture automatic processing at work³ — and crucially, the speed of a response is itself data.⁴ A choice made in 400 milliseconds and the same choice laboured over for four seconds are not telling you the same thing, even when the answer is identical.


1The fast/slow distinction is a metaphor for styles of processing, not evidence of two literal systems in the head. Kahneman (2011), Thinking, Fast and Slow (Farrar, Straus and Giroux), treats the labels as useful fictions; Evans & Stanovich (2013), “Dual-process theories of higher cognition: Advancing the debate,” Perspectives on Psychological Science 8(3), 223–241, recommend the more neutral “Type 1 / Type 2 processing.”2On the limits of introspective self-report, see Nisbett & Wilson (1977), “Telling more than we can know: Verbal reports on mental processes,” Psychological Review 84(3), 231–259.3Greenwald, McGhee & Schwartz (1998), “Measuring individual differences in implicit cognition: The Implicit Association Test,” Journal of Personality and Social Psychology 74(6), 1464–1480; improved scoring per Greenwald, Nosek & Banaji (2003), JPSP 85(2), 197–216.4On reaction time as data and its careful analysis, see Ratcliff (1993), “Methods for dealing with reaction time outliers,” Psychological Bulletin 114(3), 510–532, and Whelan (2008), “Effective analysis of reaction time data,” The Psychological Record 58, 475–482.