I have a set of measurements for web page response - 2356 6788 3658 5545 1008. How do i determine whether this set even bears evaluation / averaging. By manual inspection, this has no significance for me.
OR if I have a set like (1815 1856 1899 4567 1875), manually inspecting it i would eliminate the outlier 4567, and then use the mean of remaining values in the set.
Unfortunately due to the manual nature of data collection, increasing the size of the data set is not viable beyond 10 or so. Also, the blatant outliers like 4567 in the second set are due to known network issues.
I'm just looking for a way to plug these point into excel and answer the 2 questions above.
- Is the set statistically significant ?
- Are there any outliers that can be eliminated to make it statistically significant.