It’s often said that memories make life worth living. But in consumer insights, too often they make life a headache.
Yes, we’re talking about recall bias, which is the error that occurs when survey respondents give answers based on faulty recall. It’s a harsh reality that while memories are made to be cherished, the reason we cherish them is because they are so easily lost. And in market research, data infected by recall bias compromises everything.
As the authors of a 2017 scholarly study examining recall bias put it, “simple…recall questions generate large measurement errors.” The study focused on a survey by the Canadian government in which respondents were asked how much they had spent on food over the previous four weeks. The answers consistently underestimated actual expenditures, compared to data from subsequent diary studies involving the very same respondents (the authors noted that the diary phase was considered more reliable, but had accuracy problems of its own because of data-distorting factors such as respondents growing tired of having to record expenses each day).
The good news is there’s now an almost surefire remedy for recall bias in consumer research. But first, let’s dig a bit deeper into the problem, by asking you to take a very quick survey.