March 17, 2012 at 5:17 am #156447
Travis K. AndersonParticipant
I was wondering if someone could help me to understand what story standard deviation and variance should tell the estimator.
Std Dev: (P-O)/6
In the example below the we are calculating 1 Std Dev from the mean, i.e. the PERT value. Therefore there is a 68% certainty the estimate is 160.83 +/- 8.33 (152.5 to 169.16). At 95% the estimate is 160.83 +/- 2 * 8.33 (144.17 to 177.49).
So now what?
How do I distinguish a favorable Std Dev from a un-favorable Std Dev?
What is the Variance telling me?
How do I apply this to my final estimate?
O ML P PERT Std Dev Var 150 160 175 160.83 8.33 69.44 125 130 135 130.00 1.67
March 20, 2012 at 1:47 pm #156449
Well I typically find it useful to analyze standard deviations and variance by looking at a whole picture of data points. Two points may or may not give you a good idea. But what I notice is that as O increases so does your variance and standard deviation. This may indicate that your error measurement is not standard across all measures but has an interaction term with O (ie. there may be another variable or interaction term missing in your prediction formula).
You must be logged in to reply to this topic.