A census is a simple count or gathering of information from all people in a population [source: University of Illinois at Chicago]. An example -- with the same name, of course -- is the census taken by the U.S. government every 10 years. The first true U.S. census was conducted in 1790 and has been completed every 10 years since then. In that first census, there were nearly 4 million U.S. residents in 13 states [source: U.S. Census Bureau History]. A full census, or complete enumeration, is when every single person is counted without any estimates. But it's very difficult to conduct a full census because of problems obtaining accurate counts in remote areas, some inner-city neighborhoods and among "rare populations" that are small and not reflected in the data of standard censuses.
A sampling, or representative sample of a total population, is easier to conduct. Statisticians can extrapolate data from a sample of the full population. For example, if 10 percent of people in a sample of 150 are left-handed, we can assume that 10 percent of a population of 1,000 will be as well. Samples have a margin of error though, which gets lower as the sample size increases. In other words, sampling more people means obtaining better data.
When it comes to gathering data for market research, companies want accurate data, of course. They don't want to go to the time and expense necessary, however, to question every person in their market. That's why sampling is favored over taking a census. There are times when a census might work. If you're polling members of a small club or association -- or a small group of employees -- a census might be your best bet because you can easily track everyone down and it won't take much time to gather and compile your results. But if you want to find out how many people in the country under age 25 like your soda, it's probably best to stick with a sample of your population. In most cases, companies choose random samples -- meaning there is no bias in the group selected [source: University of California, Davis]. A stratified sample attempts to re-create the actual population on a smaller scale. For example, if you know that your town has more college graduates under age 25 than nongraduates, you can sway your sample to a similar percentage.
Are businesses becoming more socially responsible?
Answered by Jean Oelwang
How does the Iowa caucus system work?
Answered by Planet Green
Are civil unions and domestic partnerships different?
Answered by Discovery Channel