Confidence Intervals

A confidence interval is calculated to determine the probability that observed differences in two measured values may have occurred by random error alone. Typically, the experimenter sets a threshold to determine the level of confidence (p<.01; p<.o5, etc) and then calculates the resulting confidence interval for the measurements.

The Clinician's Toolkit utilizes the confidence function within Excel to display the confidence interval. The confidence function uses the standard deviation of the Benchmark Scores (residuals) and the sample size )# of cases with repeated measures) to calculated the confidence interval for the sample contained within the Toolkit.

For more information than you probably wanted to know, see: confidence intervals: wikipedia

-- JebBrown - 07 Mar 2007