2006-11-11 16:45:12 · 3 answers · asked by queen of slidell 1 in Science & Mathematics ➔ Mathematics
If the outcome of your experiment is statistical then it is not possible to tell the exact value of your result. In this case you can give an interval that will, with certain confidence, contain your result values. If, say, 95% of your values will be contained in a given interval we say this is a confidence interval containing 95% of values, or this is 95% confidence interval. If you know (or assume) statistical distribution of the result values, the confidence interval can be estimated from few result values. For instance if you have more than 30 values you can estimate avarage and variance and use Gaussian distribution to obtain confidence interval. For less than 30 values the student distribution is used.
2006-11-11 18:11:02 · answer #1 · answered by fernando_007 6 · 0⤊ 0⤋
THis is the range (from 1 to 5 OR from -.22 to +.67) that a statistical result exists. Let's see: If there is a study that asks how many times does the average person go to the bathroom in one day....in the USA? The result is a bunch of numbers anywhere from 0 to say 10 times (whoa!). So after doing some statistical analysis.... the mean numer is 5. But there is a level of error invoved in ANY study. So, the confidence interval is from 3.3 times to 6.7 times. And Confidence Intervals are always associated with a percent....95%, 99%.....95.7%.
2006-11-12 00:54:51 · answer #2 · answered by spidy 2 · 1⤊ 0⤋
Confidentially said, Spidy's answer is accurate, simple, pleasant and... speedy!
2006-11-12 01:21:40 · answer #3 · answered by 11:11 3 · 0⤊ 0⤋