Now you have a winner of the
taste test (really, more of a survey) consider the following
question. If you were to announce the winner to the world, what
would you say?
Would the headlines be:
"The superior beverage in the U.S. is ... (insert
"The favorite drink of the people we surveyed is ...
(insert your winner)"
Aha! Do you have any right to claim that the people of the
United States all love the winning beverage?
Then, how do manufacturers and political spinsters, during elections,
claim that a product or candidate is preferred based on only the
opinions of thousands or fewer people?
The answer is, those who make the claims, do so because
they feel their findings are significant. To understand what is
significant we must look at statistics.
The reason your findings, as raw data, are still not
significant is that you have not sufficiently ruled out the possibility
that your result is possible by chance, i.e. a freak occurrence, fluke,
or random event. You need to prove to a reasonable level that this is
For an example of how to do this go to the next