 |
|
 |
|
|
|
|
|
|
|
|
||
|
|
|||||||
|
|||||||
|
|
|||||||
|
|||||||
|
|
Asking the right questions — attitude specificity
One reason that attitudes sometimes fail to predict behavior can be traced to the way attitudes and behaviors are measured. Often, attitude surveys ask fairly general questions. What do you think about birth control? Would you describe yourself as a liberal or conservative? But behaviors are measured at very specific levels. Do you use the pill? Did you vote for a particular liberal or conservative candidate? Asking a general attitude to predict a specific behavior is a pretty tall order. The decision to use a particular form of birth control, for example, or to vote for a particular candidate, depends on several attitudes. Will this form of birth control also protect against infectious disease? How does this candidate feel about the environment? The point is that general attitudes may predict general behaviors (e.g., are you using any form of birth control), but they won't do very well at predicting specific behaviors <REF>(Fishbein & Ajzen, 1974). On the other hand, specific attitudes do very well at predicting specific behaviors. Election polls, where respondents are asked whom they would vote for if the election were held tomorrow, are accurate within a 3% margin of error. Attitudes are also more likely to predict behavior when they are formed through direct experience with the attitude object <REF>(Fazio, 1990) and when they are in areas that you consider relevant to your self-concept and interests <REF>(Boninger, Krosnick, & Berent, 1995). One of the interesting things about attitudes is that we can generate them on the fly. For example, how would you feel about Washington D.C. being made the 51st state? Although you may never have thought about statehood for D.C., you can probably generate an attitude about it. But, unless you live in D.C., that attitude won't be very predictive of your behavior because it is not very relevant to your life. |
||||||||||||
|
