Confirmation bias
Confirmation bias is the tendency to seek out, interpret and remember information in ways that confirm what we already believe. In research it leads teams to hear the evidence for their idea and to discount the evidence against it.
How it works
The bias works at every stage of a study. It shapes which questions get asked (leading questions that invite agreement), whom the team recruits (customers who already like the product), what analysts notice in transcripts and which quotes reach the final report. It operates below awareness, so knowing about it is not enough to prevent it.
The psychologist Peter Wason's experiments in the 1960s showed people testing rules by looking only for confirming cases. Research teams counter it with neutral question wording, samples that include critics and former customers, more than one analyst, and a deliberate search for evidence that would disprove the hypothesis.
A worked example
A retailer's product team believes shoppers want a loyalty app. Its first 10 interviews are with top-tier members, and 9 say yes. A second round with 20 occasional shoppers, the people the app is meant to win, finds 14 of them would not download another store app at all.
Related
Automate this in Pulse
Pulse's themes are built from every interview in a study, the awkward ones included, which hands a team with a favorite hypothesis the counter-evidence in participants' own words.