Saturday, November 12, 2005
Mean and Standard Deviation Calculation
I was reading up on PERT (Program Evaluation Review Technique), one of the technqiues in estimating efforts in WBS (work breakdown structure). It appears that there is much more to estimation than just ball-park guesses.
I will share my lessons learnt on PERT later; however it appears the following are relevant lessons that needs to be learnt, as they are applied in PERT calculations.
I will share my lessons learnt on PERT later; however it appears the following are relevant lessons that needs to be learnt, as they are applied in PERT calculations.
Standard deviations is very important in PM estimations. One of the usage is in knowing the risk level of estimates. For example, I have the following sample of WBS:
Work Estimated Effort Std Deviations
Module A 20 man-days 1.0
Module B 20 man-days 2.0
What this means is that although both tasks are estimated to take 20 man-days to complete. Module B has a higher risk level as the higher deviation means that the actual effort is in the range of 20 man-days +/- 96%, as compared to Module A's 65%.
Work Estimated Effort Std Deviations
Module A 20 man-days 1.0
Module B 20 man-days 2.0
What this means is that although both tasks are estimated to take 20 man-days to complete. Module B has a higher risk level as the higher deviation means that the actual effort is in the range of 20 man-days +/- 96%, as compared to Module A's 65%.
