STAT 301 Lecture Notes - Lecture 13: Confidence Interval, Solitary Confinement, Central Limit Theorem
Document Summary
Estimating a difference in means - part 2. In the previous examples, we had two groups of data whose means were comparing: prisoners in solitary confinement vs those in regular cells, tomato plants receiving fertilizer a vs. fertilizer b. Sometimes, instead of two separate groups of subjects, we have data for which each subject has been measured twice, and we want to draw inference on the average difference btween the two observations on the same subjects. "paired" data: before and after type studies (e. g. compare blood pressure before being put on a drug to after being on the drug) Studies where each subject is measured under twto different treatments, or a treatement and a control: when we measure each subject twice, we call the two observations "paired" observations. If we want to draw inference using differences between paired observations, we can just subtract the paired observations from each other and make a confidence interval for the mean of the differences.