Related Web Sites

Sampling in Research
A tutorial on creating samples to obtain accurate research data.

The Survey: Sampling, Questionnaire, Distribution and Interviewing
This series of lecture slides, from the New Jersey Institute of Technology, provides a simple overview of the theory and methods of random sampling.

What Is a Random Sample?

What's Blood Got to Do with It?
A blood test functions as a good example of a random sample. If you go to the hospital emergency room with severe abdominal pain and a high fever, the doctors might suspect that your appendix is causing the pain. However, before they pull out the scalpel and start cutting, they, thankfully, take measures to confirm their suspicion.

A person suffering from appendicitis has significantly more white blood cells than normal. To discover your white blood count, do the doctors drain out all your blood and count the cells? Of course not! They prick your finger with a needle, remove a drop of your blood, place it under a microscope and count the white blood cells. If there are twice as many white blood cells in that drop of blood than is normal, the doctors can predict that the white blood cell count is higher than normal throughout all your blood.

Being Certain It's Random
The registered voters in our fictitious mayoral race are like a bloodstream, with one important difference. If we prick the skin on any part of your body to draw out a drop of blood, we can be certain that that drop has the same properties as the rest of your blood. If, however, we "prick the skin" of the registered voters in a section of the city where members of only one ethnic group with average yearly incomes above $250,000 live, the "drop" we get will not be representative of all the registered voters in the city. This is why it is necessary for pollsters to create random samples that accurately represent a cross section of the entire population.

[+/-5%]