You saw in Chapter 10 how to introduce correlation into an @RISK simulation with RISKCORRMAT functions. We implicitly assumed that the activity durations in Example 15.5 are probabilistically independent.

However, it is very possible that some of them would be correlated in a real situation.

Specifically, assume activities A and B are positively correlated with correlation 0.7. Also, assume that activities G and J are positively correlated with correlation 0.6. Modify the model appropriately and rerun the simulation. What differences, if any, do you see in the outputs?