"What positive future events does this action cause?"

When reading this, Thomas Sowell's 3 questions came to mind.

• Compared to what?
• At what cost?
• What data do we have?

Without identifying the other options, even identifying something as "the cause" becomes problematic. And it is always problematic because "the cause" is generally used as "that event which I assign the credit or blame to".

If you're only looking for the positive events, you're obviously biasing your search. You should be looking for all consequences first, good, bad, and neutral. Further, by encouraging people to start looking for either the good or bad, they're starting off with a position and attitude first, then the reality of events second.

Finally, you have to consider the data you have to back up your claims of the likely consequences. What's your confidence in the model?

Checking consequentialism should answer Sowell's three basic questions -

• Compared to what? ** What are my options?
• What data do you have? ** What probabilities do I assign to possible states of the universe for each state?
• At what cost? ** What value do I assign to these different states?

(What am I wrong with the asterisks? I wanted my versions of each of his questions to have an extra level of nesting.)

All these things come out automatically if you follow the mathematical formalism for decision theory. That would be my suggestion - have people work out an everyday problem in the formalism of probabilistic decision theory. Show them that the problem has already been solved, and they just have to turn the crank.