This

**preview**shows half of the first page. to view the full**2 pages of the document.**INST354 Lecture 12: Probability

We will use probabilities (in the range from 0 to 1) to represent beliefs about what will happen.

Usually we mean to summarize people’s subjective beliefs about those events. Although we use

numbers that might be interpreted as formal probabilities by a mathematician, we do not

assume that these numbers necessarily behave like true probabilities. In fact, one of the

important discoveries of psychological research is that subjective probabilities are not always

consistent with mathematical probabilities. (Chapters 7 and 8 summarize many of the ways in

which our judgments under uncertainty violate rules of formal probability theory.) When we

mean to refer to mathematical probabilities, we will make sure the context is clear. (The

Appendix in this book introduces the mathematical laws of probability.)

We will not spend much time in this book on how these numbers summarizing consequence

values and outcome uncertainties might be extracted from people’s thoughts about decision

situations, but psychologists and economists have developed many useful scaling methods to

solve these measurement problems. To spare the reader a lot of technical detail, we will usually

just present plausible numbers. The reader who wants to understand these methods can find

this information in many other sources (e.g., Dawes & Smith, 1985)

.

We will often use simple gambles to illustrate decision-making principles and habits. Gambles

are the most popular experimental stimulus in research on decision making, and they provide

well-defined, easy-to-understand decision dilemmas in situations where we can be sure that our

research participants want to “maximize” the amount of money they earn in the experiment. So,

let’s work through the representation of a typical experimental gamble in terms of the decision

tree diagrams. Consider the choice between two gambles we described in Chapter 1:

Figure 2.2 summarizes this situation in a decision tree diagram—when the outcomes are

naturally scaled with meaningful numbers like dollar amounts, we will just use those numbers for

clarity (rather than the 0–100 scale we use for more subjective outcomes). An interesting

question, which is of practical importance for judgment researchers, concerns the extent to

which human thinking is the same both in crisp, well-defined gambles and in ambiguous

everyday situations (like the knee operation; Lopes, 1994, provides a thoughtful discussion of

###### You're Reading a Preview

Unlock to view full version