have the right answer. It would be wrong, however, to infer from that answer that you are incapable of
getting to the restaurant. If given more time or a sufficient incentive, you may change your response. You
may have a map to reference, or a spouse to ask, to find the correct answer. While you might not walk
around with the directions to the restaurant in your head, you have a place in your life where such
information is stored and available for quick retrieval when you need it.
The difference between the facts we can recite in the context of an unexpected and fast-paced survey
interview readily and the knowledge we need to accomplish a larger range of less-frequent tasks is why
we keep books on our shelves, the phone numbers of trusted friends in an easily accessible place, and
computers on our desks. Knowing where these references expands the knowledge base that is in effect for
most of the decisions we make. These references help us organize information for quick retrieval that we
can use when we have the motivation and opportunity to do so. Most political surveys offer neither the
opportunity nor the motivation to consult such references.
We contend that people may perform poorly on political knowledge tests not because they are
incapable of answering the questions, but because they are caught unprepared and unmotivated to do well.
If true, existing political knowledge scales offer biased and unreliable measures of what people know
when they vote and make other important political decisions.
We examine how changes in motivation and research opportunity affect responses to political
knowledge questions. Our approach is experimental. We vary two factors -- time and reward -- within a
set of survey interviews. We give one randomly selected half of our sample only one minute to answer
each question, whereas the other half has 24 hours to come up with the answers. Independently, we offer
one randomly selected half of our sample a small monetary reward for each correct answer. Since the
variations are independent, each respondent is equally likely to be in one of four conditions: one minute
no pay; one minute with pay, 24 hours no pay, 24 hours with pay. These variations are important because
the standard way of measuring political knowledge is to place all respondents in the “one minute no pay”
context. Henceforth, we refer to these respondents as the control group.
2
Convention