John Carpenter runs a timber company. John is considering an expansion to his product line by manufacturing a new product, garden sheds. He would need to construct either a large new plant to manufacture the sheds, or a small plant. He decides that it is equally likely that the market for this product would be favourable or unfavourable. Given a favourable market he expects a profit of $200,000 if he builds a large plant, or $100,000 from a small plant. An unfavourable market would lead to losses of $180,000 or $20,000 from a large or small plant respectively.
(a) Construct a payoff matrix for John's problem. If John were to follow the EMV criterion, show calculations to indicate what should he do, and why?
(b) What is the expected value of perfect information and explain the reason for such a calculation?
John has the option of conducting a market research survey for a cost of $10,000. He has learned that of all new favourably marketed products, market surveys were positive 70% of the time but falsely predicted negative results 30% of the time. When there was actually an unfavourable market, however, 80% of surveys correctly predicted negative results while 20% of surveys incorrectly predicted positive results.
(c) Using the market research experience, calculate the revised probabilities of a favourable and an unfavourable market for John's product given positive and negative survey predictions.
(d) Based on these revised probabilities what should John do? Support your answer with EVSI and ENGSI calculations.
(e) The decision making literature mostly adopts a rational approach. However, Tversky and Kahneman (T&K) (Reading 3.1) adopt a different approach, arguing that often people use other methods to make decisions, relying on heuristics.
What do they mean by the term heuristics?
Describe three types of heuristics that T&K suggest people use in judgments under uncertainty.
Give one example from your own experience of a bias that might result from each of these heuristics.