Wednesday, November 19, 2008

Confirmation bias in market research: The circular references

If you've ever implemented a formula in a spreadsheet like Excel, you've probably experienced the Circular Reference Warning at some time or another. That's the one that tells you your formula relies on the value of itself to calculate itself...a self-referencing system of sorts.

I was reminded of the self-referencing nature of some market research when participating in an online survey panel recently. The survey contained many questions with four-box response options. For example:

How likely would you be to invest in a company whose reputation is environmentally friendly?

1. Highly unlikely
2. Somewhat unlikely
3. Somewhat likely
4. Highly likely

Overlooking the construct of the question itself, the challenge with this response format is that it forces the respondent to have an opinion..What if I am neither? Certainly a four box response provides a neat and tidy interpretation for the researchers...no ambiguity, no fence sitting. But does it capture any truth?

What's wrong with discovering that your survey audience has no opinion? Isn't that an opportunity for marketing? Or would we rather that the survey force the issue? Requiring that the audience declare themselves as conforming to one side or the other of a binary world view?

If the opposite of love is apathy how does a four-box, love-hate response provide a reference point we don't already possess?

No comments:

Post a Comment