STAT1008 Lecture Notes - Lecture 13: Confidence Interval, Statistic
Document Summary
Example: mosquito likelihood to approach humans after being infected with malaria virus during first 8 days (before become infectious) and during infectious period. Pc: proportion of mosquitoes in the control group to approach human. Pe: proportion of mosquitos in the exposed group to approach humans. If a randomisation distribution is normally distributed we can write it as: When doing the above in reverse to find statistic use either negative or positive p-value depending on whether the result should be greater or less than null value. Creating confidence intervals: bootstrap confidence interval, using normal distribution- same as bootstrap but with smooth curve, finding the z* value on a standard normal distribution and then using the formula statistic +/- z*. Se: x = mean +/- z. sd (comes from z formula, making x the subject to unstandardized)