I have a question: In the Multistage problem given for practice near the end of this video, I see the final decision was only about choosing the first stage decision ( i.e. choosing of Investment A due its Higher EMV value than that of B ). Do we stop at that point? Or Should the final answer also include, the decision at the 2nd stage? ( i.e. after choosing Investment A, then we arrive at decision node 4. Which one to choose at that node? In short: Should the answer also speak about the consequent decisions after the main 1st stage decision.?
That's a good question! My understanding is they are estimates about the accuracy of the test market study. People always leave out where they get these from in the textbooks!!!. Where do you get the probabilities? I think it would come from historical data. You look at past studies and see which demand conditions materialized after the historical studies. In this case, 9.8% of the time, low demand materialized after a favorable study, 39.4% of the time Medium demand materialized after a favorable study, and .507% of the time, high demand materialized after a favorable study. These add up to one. You need history to do this. My hope would be that a good market research firm would have these probabilities available if they do these kinds of studies all the time. If they couldn't provide such values, I'd be pretty disappointed, and I'd wonder how many times they'd actually done such a study. If picking stocks based on recommendations from analysts, it might be easier to come up with the probabilities because they often publish their recommendations, and you can see whether their recommendations actually panned out. You could use this data to generate probabilities for say, low, medium and high returns based on favorable (buy) or unfavorable (sell) recommendations.
Wonderful lesson, still reasonating easy and simple 6years later. Thank you for all you do.
Wonderful
I have a question: In the Multistage problem given for practice near the end of this video, I see the final decision was only about choosing the first stage decision ( i.e. choosing of Investment A due its Higher EMV value than that of B ). Do we stop at that point? Or Should the final answer also include, the decision at the 2nd stage? ( i.e. after choosing Investment A, then we arrive at decision node 4. Which one to choose at that node?
In short: Should the answer also speak about the consequent decisions after the main 1st stage decision.?
nicely explained
Where did the probabilities come from?
That's a good question! My understanding is they are estimates about the accuracy of the test market study. People always leave out where they get these from in the textbooks!!!. Where do you get the probabilities? I think it would come from historical data. You look at past studies and see which demand conditions materialized after the historical studies. In this case, 9.8% of the time, low demand materialized after a favorable study, 39.4% of the time Medium demand materialized after a favorable study, and .507% of the time, high demand materialized after a favorable study. These add up to one. You need history to do this. My hope would be that a good market research firm would have these probabilities available if they do these kinds of studies all the time. If they couldn't provide such values, I'd be pretty disappointed, and I'd wonder how many times they'd actually done such a study. If picking stocks based on recommendations from analysts, it might be easier to come up with the probabilities because they often publish their recommendations, and you can see whether their recommendations actually panned out. You could use this data to generate probabilities for say, low, medium and high returns based on favorable (buy) or unfavorable (sell) recommendations.