TCTTS said:
"Good to know" that the conclusion reached from the data could be based on bad analysis, even thought "I have no idea" what multivariate statistics are.
No worries. When I get a chance I'll type up why their analysis was flawed. Likely won't be until this weekend as I'm in the middle of a job transition.
In short, they have a few arbitrary variables in there that they failed to show any kind of sensitivity analysis on. They applied the formulas "correctly", but not in a way that actually allows for any meaningful conclusions because of how they defined them and because of the variables they left out.