The concept of confidence intervals is general. You can calculate the 95% CI for almost any value you compute when you analyze data. We've already discussed the CI of a SD. Other confidence intervals computed by Prism include:

• The difference between two group means

•A proportion

•The ratio of two proportions

•The best-fit slope of linear regression

•The best-fit value of an EC50 determined by nonlinear regression

•The ratio of the median survival times of two groups

•The median of a set of values.

The concept is the same for all these cases. You collected data from a small sample and analyzed the data. The values you compute are 100% correct for that sample, but are affected by random scatter. A confidence interval tells you how precisely you have determined that value. Given certain assumptions (which we list with each analysis later in this book), you can be 95% sure that the 95% CI contains the true (population) value.

The fundamental idea of statistics is to analyze a sample of data, and make quantitative inferences about the population from which the data were sampled. Confidence intervals are the most straightforward way to do this.