MANAGERIAL DECISION MAKING

D.  The Value of Information

1.  Expected value of information

.

.

.

.

.

.

• Suppose perfect information

.

.

.

.

.

.

.

• Expected value of information (EVI) - difference between expected value with information and without it

.

.

• Acquire information if EVI > cost of information

2.  Imperfect information

• Possible that mistakes are made

.

.

.

.

.

• Prior probability - probability of an event before new information is obtained

• Posterior (conditional) probability - probability of an event after new information is obtained

.

.

.

.

.

.

.

.

.

.

.

.

3.  Bayes' Theorem

.

.

.

Ex. -

.

.

.

.

.

.

.

.

Ex. -

.

.

.

.

.

.

.

.

.

.

.

.

.

4.  Intuitive prediction

• Prediction based on personal judgment and experience

• Problems:

(1)  Difficult to examine logic underlying prediction

(2)  Forecasts frequently inaccurate or biased

(3)  Individuals tend to be overconfident in ability to make predictions

(4)  Too much weight put on new information, not enough on old information (i.e., prior probability)

(5)  Not enough weight given to statistical information

4.  Optimal stopping

• If an initial investment is unsuccessful, how much further investment should be made?

a.  Increasing probability of success

.

.

.

.

.

.

.

.

b.  Decreasing probability of success

.

.

.

.

.

.

.

5.  Optimal sequential decisions

• In what order should sequential decisions made?

.

.

.

.

.

.

• Invest as long as p > c / predicted profit

• Sequence of investment should be in descending order of p / c

Ex. -

.

.

.

.

6.  Simultaneous search

• Increase alternatives at a cost

.

.

.

.

.

.

.

.