The difference is quite a bit. Here are Gallup's current poll results:
According to this poll, Romney leads among "registered voters" by a mere 1% (48 vs. 47). His lead among "likely voters," however, is a strong 7% ( 52 vs. 45). In other words, the difference between "registered voters" and "likely voters" can be, in this case, the difference between a statistical tie (48 vs. 47) and a landslide (52 vs 45).
There is a fairly consistent trend: "registered voters" are to the left of "likely voters." Thus, the less interest one has in politics, as evidenced by not voting even though registered, the more likely that that person is to have liberal opinions.
This large difference is also one reason by polls are unreliable as a predictor of election outcomes. It is difficult enough for pollsters to reach a large and representative sample of adults and get them to answer the survey. The pollster also has to estimate (guess) how likely these respondents are to vote in an upcoming election.
No comments:
Post a Comment