Class Notes (1,100,000)
US (450,000)
UMD (8,000)
INST (100)
Anton (10)
Lecture 12

INST 354 Lecture 12: INST354 Lecture 12: ProbabilityExam


Department
Information Studies
Course Code
INST 354
Professor
Anton
Lecture
12

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 diagramwhen the outcomes are
naturally scaled with meaningful numbers like dollar amounts, we will just use those numbers for
clarity (rather than the 0100 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