Five pilot testers report times of 51-56-83-100-150.The three closest grouped times are taken as representative; that is, 51-56-83. These total 190 minutes which divided by 3 equals an average of 63 minutes. 63 divided by 50 yields 1.0 CPE (rounded to the next lowest hour). If all five times were used the CPE would be 1.5 which would be distorted by the unrealistically high time of 150.
Five pilot testers report times of 37-48-49-90-90. If we chose 48-49-90 we get 1.25 CPE rounded down to 1.0 CPE. If we do 37-48-49 we do not get even 1 CPE. If we do 49-90-90 we get 1.5 CPE. The most realistic choice is 48-49-90 yielding 1.0 CPE or to use all five times which again results in 1.0 CPE.
Six pilot testers report times of 32-55-63-66-72-120. If we remove the low and high outlying times we have a fairly good grouping of times 55-63-66-72 which yields 1.0 CPE. If we use all times we still arrive at 1.0 CPE. If we removed the low outlier we would compute 1.5 CPE which would distort the number.