A very common design of repeatability study it to take just two measurements on each person. The results of such as study on five people are illustrated in the following figure (download Excel spreadsheet here).

To calculate the SEM we proceed in exactly the same way as we did for the previous study. Because there are only two measurements for each person part of this requires us to calculate the standard deviation of those two measurements. Although we would very rarely take the standard deviation of just two measurements in any other circumstances this is still valid and you should find that the corresponding spreadsheet functions still work.

As in the previous example the root mean square average is used to calculate the overall SEM {=√((5.7²+4.2²+5.7²+2.8²+2.1²)/5)}.

Next page: Calculating SEM using Analysis of Variance (ANOVA)