ABSTRACT
The other calculations in this chapter involve the frequency of response in the control and test groups of experimental studies. Once the test and control group response frequencies have been calculated, the difference between them is assessed for statistical significance. An experimental study is most useful if the difference between the frequency of response in test and control groups is statistically significant. Also, the test group response frequency must be adjusted for the frequency of response in the control group. Finally, if there is sufficient reason (e.g., small sample size, nonoptimal control group, short followup time), the upper confidence limit of the control-adjusted response frequency should be computed and used in risk assessment.
