We encounter many issues in our daily life that depend on statistics:
Can a nation-wide poll of 200 voters show that candidate Joe will be elected President of the U.S.?
Can a survey of 10 gas stations show that gas prices are rising?
Can one determine the effectiveness of a certain medication using a study with 70 patients?
Answering these questions requires that we understand how small samples are error-prone, and how we determine the magnitude of the error.
We will explain the issues that arise when dealing with statistical phenomena through the example of coin tossing. One believes that coin tossing provides equal probability of getting "heads" or "tails".
Can we prove this?
Try tossing a coin 10 times. Sometimes you will get five heads (or tails), but often you will get other values - 3,4,6, or 7, and less frequently 2 or 8. Rarely, you will get 0, 1, 9 or 10.
Try it: Click the "Toss" button
again several times, and then use the pulldown box to try different numbers of
Last Modification - July 11, 2004