Many tests are designed to measure some aspect of the population - such as a difference between means of two groups, or a relative risk - based on a sample.
The reported value is the best estimate, and is subject to random sampling error. The confidence interval gives is a range likely to contain the true value of the parameter.
The 'level of confidence' can vary: a 95% confidence interval will contain the true value 95% of the time. If a higher level of confidence is required, the interval must be wider.
Confidence intervals are constructed using statistical techniques that analyse the way random sampling errors affect the estimates given. Biostats Calculator will automatically generate confidence intervals for many of the values it calculates.
Biostats Calculator performs tests integrating confidence intervals, and several other calculations from the convenience of your iPhone or iPod Touch. Buy it today!
