PERT Estimation (Program Evaluation and Review Technique)
PERT or Program Evaluation and Review Technique is used for time (activity duration) and cost estimation.
PERT calculation is a special form of three-point estimation which uses optimistic, most likely and pessimistic values for cost and duration estimation.
Besides giving the expected value as the result of weighted calculation, pessimistic and optimistic values enables the project managers to calculate the standard deviation and variance.
What is PERT according to PMI?
PMI defines Program Evaluation and Review Technique aka PERT as a technique for estimating that applies a weighted average of optimistic, pessimistic, and most likely estimates when there is uncertainty with the individual activity estimates within PMBOK 5th Edition page 553.
PERT (Program Evaluation and Review Technique) is a special form of Three-Point Estimating technique which is used as a tool or technique applied at Estimate Activity Durations process in Schedule Management Knowledge Area.
For more about Three-Point Estimating, PMP certification candidates can refer to PMBOK 5th Edition page 170 (188.8.131.52 Three-Point Estimating).
As PERT is a weighted average of three estimates namely as optimistic, most likely and pessimistic the formula which weights these three different estimates can be described as follows:
tE = (tO + 4*tM + tP) / 6
tE Expected: tE is the Expected value as a result of PERT Program Evaluation and Review Technique
tM Most likely: tM is the most likely estimate which is the estimate that is expected to happen for an activity duration.
tO Optimistic: tO is the minimum estimated activity duration as a result of the best-case scenario for an activity.
tP Pessimistic: tP is the pessimistic estimate or the highest value estimated for an activity duration on the worst-case scenario.
Standard deviation can be calculated as follows:
standard deviation = (tP - tO) / 6
And the variance is calculated using the below PERT formula:
variance = [ (tP - tO) / 6 ]^2
As a project manager, if you are making an estimation for a number of activities note that the standard deviation can not be summed.
For example, if you have two activities namely activity A and activity B, the overall standard deviation equals to:
Standard Deviation(of A and B) = SquareRoot of [ Variance(of A) + Variance(of B) ]