The actual type of poll you are referring to in this
hypothetical is called a "straw poll", and in polling organizations it is generally
believed to be the least reliable kind of survey there is. We see these kinds of polls
online and in magazines all of the time.
On a sports
website, for example, they may have a poll to see who you think is going to win the
Superbowl this year. The results, just like in the survey you mention, cannot
accurately measure the population in any meaningful way. It is based only on those who
happen to see or read the questionnaire or poll, those who choose to fill it out, and
then those who actually choose to mail it in. This is not going to be a representative
group of people. Even if it were, as pohnpei points out, you wouldn't know who or what
they represent, because no other information about the respondents was gathered. Was it
72% of all who responded? Of women? Senior citizens? The details of such demographics
matter when determining how accurate a poll can be.
No comments:
Post a Comment