An ever important subject, but like most statistical analysis tools often misunderstood and not used correctly!

Will spend more time on pitfalls in interpretation of data on the topic of statistic!

Lets start with  a simple but important tool.

The Cost Performance Index. or acronym CPI

I bring this up in risk analysis for whenever managers owners describe to me their shop or project they point out the size of the shop or the number of employees. In another words their expenses. But nobody tells me their Cost Performance Index which is true measure of the efficiency of the project or business,  In another words how much return on the dollar on project or business one gets. The efficiency of this is also a starting point as to how much risk one can assume!

     Here is the definition of the CPI!     formula                                   


A CPI of greater than one indicates cost below budget estimates, a CPI of less than one indicates cost overrun of estimates. Now lets get into some details. As to what that means! To start with that ratio is based on historical data,It allows one to compare an old project to new project that is similar.

One can also do an analysis as one progresses on the project see how it compares to expectations!When used like this there is lots of conjecture and interpretation involved in the data. As always keeping detailed records of tasks and ones Work Break Down Structure is essential.

Included here I have an article on Probabilistic Risk Analysis Foundation and Methods by Tim Bedford and Robert Cook.

It sites  a General Electric full numerical probabilistic risk assessment the likely hood of a successful landing on the moon a chance of success was less than 5%.

But yet NASA made many successful landings on the moon! Are the numbers wrong? No whenever one sees statistical data like this use it as a guide not as absolute.

What that means is the higher the risk pay more attention to detail, in complex situations the risk is not equally disbursed.

In application knowing the failure rate of product or part helps immensely in mitigating loss. So staying informed on research is essential and knowledge of physical characteristics of the situation.

There are two tools that are used to asses the risk prior to mathematical analysis if so desired!

One is called the Risk Matrix,I find it useful in general initial discussions with clients, but not much use in the field. The other is a Failure tree analysis, what it means is assessing a situation that would result in complete failure given certain conditions.