Question on correct way of calulating the standard error of the mean?!
Question: Question on correct way of calulating the standard error of the mean?
I have carried out three repeats of an experiment each containing three readings and need to present the data showing the standard error of the mean (SEM). Which is more statistically correct, using all the raw data, i.e. nine datapoints, together or the three individual means to calculate the SEM?
Answers:
The two ways are equivalent, since one of the theorems of statistics is that: The sum of the means equals the mean of the sum. Averaging all nine at once is at least briefer in its number of steps.