SPERT-Beta Development Release ‘A’ available

Well, this past week instead of blogging, I spent a huge chunk of my time working on the next edition of Statistical PERT:  Statistical PERT – Beta, which will use MS Excel’s built-in beta distribution functions to easily create probabilistic estimates with greater accuracy and for a greater number of uncertainties, even those that are too skewed to use Statistical PERT built with the normal distribution functions.

I’m really excited about this next edition of Statistical PERT.  This next edition will be a companion to, not a replacement for, the original edition of Statistical PERT using the normal distribution.  Here’s how I characterize the two editions:

Statistical PERT – Normal

  • Very easy to use
  • Very easy to modify
  • Very flexible
  • Easy to understand and explain
  • Can easily use the SPERT concepts without a pre-built SPERT template
  • Works with any bell-shaped uncertainty
  • Works with slightly-to-moderately skewed uncertainties
  • Accuracy is between 95% – 99.9%

Statistical PERT – Beta

  • Very easy to use
  • Not as easy to modify
  • Very flexible
  • Not as easy to understand or explain
  • Cannot easily use the SPERT concepts without a pre-built SPERT template
  • Works with any bell-shaped uncertainty
  • Works with even very severely skewed uncertainties
  • Works with non-bell-shaped uncertainties
  • Achieves very high accuracy (not sure exactly how accurate yet, but it is likely that accuracy will be > 99% for the uncertainties it is designed to estimate)

Want to learn more about the next edition of Statistical PERT?  Visit the new Statistical PERT Beta Blog to learn more and download a development release of the SPERT-Beta Excel workbook.  SPERT-Beta will only be available for downloading on the Statistical PERT Beta Blog until the stable Version 1 is released (this coming winter, 2016).

Leave a Reply

Your email address will not be published. Required fields are marked *