POPM 3240 Lecture Notes - Lecture 5: Central Limit Theorem, Sampling Distribution, Standard Deviation
Document Summary
Why do stats and epi get along? in epi, we may want to estimate how commonly something occurs or we may want to explore differences in risk of disease between groups. Need to demonstrate statistical association (among other things) to know whether observed differences are real . Statistical analysis = using the right tools & interpreting the results correctly. On average (if samples are representititve) we should get the right answer. To calculate: add all individual values in the group and divide by the number of values in the group. Examples include blood test results (normal ranges), weights, incomes (average income), scales (scales of agreement) Limitations include that it doesn"t show distribution (could be wide or narrow) Would have a variance = s2, which describes the overall variability around the mean (all directions) Would have a standard deviation (s) = (s2)1/2 , which describes the variability of the data in one direction. This is the basis of the z-test .