Tuesday, July 16, 2013

unincentivized surveys

This was obviously written by an economist :) I would have emphasized the final paragraph rather than the first one. Psychologists definitely rely on unincentivized surveys too much, but economists are too distrustful of unincentivized questions, because they don't distinguish between situations in which biased answers are likely from those in which they aren't. It's not always possible or practical to incentivize survey questions, and if you don't have a reason to doubt people's responses, why not accept them at face value for the time being?

Experimental economics started off focusing only on the domains in which incentivization definitely matters, such as market activities. But now behavioral economics is invading psychology, and it's not always so important to incentivize correctness when you're just trying to directly measure thought processes. That is, if you ask someone what they think about an economic activity, they might tell you something that bears little resemblance to their actions when participating in that activity. But you're quite likely to hear something that resembles what they actually think about the activity...

No comments: