• One of the basic premises of Buss' paper can be found on p. 206: parental investment and its consequences for sexual selection. No doubt, there are biological differences between men and women that may lead to asymmetries in parental investment. Paper topic: What social and cultural processes exacerbate or eliminate these asymmetries?
• Table 1 summarizes all hypothesized gender differences. The rest of the paper provides a more detailed discussion of these differences and empirical data pertinent to them. Paper topic: Pick one study that convincingly confirms a prediction and one study that does not. What are the differences between the two cases? Paper topic: Pick two studies and try to explain the results in terms of social-cultural pressures rather than evolutionary strategies.
• Paper topic: From the standpoint of the study of decision making, what is the relationship between Buss' "strategies" of mate selection and actual decisions of selecting a mate? (Many answers possible; think about conscious versus unconscious choice; disposition vs. actual behavior; multiple determining factors, etc.)

ch 2: Cognitive dissonance

• Read carefully the description of the famous 1959 experiment (p. 23-24). What exactly causes the cognitive dissonance? The fact that one has done something odd (or boring, or embarrassing...) for no good reason might be crucial. Our society demands that our actions have rational (and acceptable) reasons; if we do things that don't sem to have a good reason, we sometimes have to make up one... This key element in dissonance theory relates to the article "reason-based choice" and to the discussion of the "folk-psychological perspective" on decision making. Paper topic: Discuss the relation between dissonance processes, reason-based choice, and the folk psychology of decision making.
• Note the separate discussion of pre-decision dissonance and post-decision dissonance. What causes the one, what causes the other? Plous discusses commitment as one form of post-decisional dissonance reduction. In lecture, we discussed other forms as well.

• Shafir et al. have collected a series of studies that fit only moderately well together. The paper lacks theoretical rigor, but the results of the studies are so interesting that it is worth reading them carefully.
• In particular, I invite you to come up with several alternative explanations to each study. Shafir et al. favor the "irrationality" frame of interpreting the results. How could we make sense of people's behaviors in the studies without assuming that they acted irrationally?
• In class, we discussed the anticipatory function of decisions. In making a decision, people try to anticipate the consequences of their choice. Part of these consequences are feelings of regret about one's own decision. Could some results in Shafir et al.'s collection be explained by people's tendency to anticipate their own regret and therefore to choose the option that least likely produces regret?

• EUT (= expected utility theory) suggests an elegant solution to the problem of choosing between options that differ in their consequences. For each option, you calculate a "utility," which is this option's desirability multiplied by the likelihood of attaining it. When considering whether you should ask your neighbor out on a date, you take into account how much fun it would be to have a date with her or him (= desirability) and how likely it is that she or he says Yes in the first place.
• More precisely, in evaluating each option, you try to come up with all its possible consequences and then sum, across all consequences, their desirabilities multiplied with the respective likelihoods. The number at which you thus arrive is called "expected value" or "expected utility."
• Note that representing and anticipating future consequences of an option is an essential ingredient of EUT. Also, as pointed out in Malle (1995), desirabilities correspond to the folk-psychological concept of desire, and likelihoods correspond to the folk-psychological concept of belief.
• EUT makes several assumptions. One assumption is that utility can be measured as a function of wealth (i.e., money); several other assumptions concern the mathematic operations of deriving and combining consequences of your options (see p. 81). The thrust of EUT is that if a person makes decisions that fit with those assumptions, then those decisions can be described as maximizing utility. Because some of these assumptions (and the goal of systematically maximizing one's utility) have been associated with the concept of rationality, EUT is sometimes called rational choice theory
• What do we mean when we say that a decision was "rational"? How does rationality relate to the assumptions of EUT?

• EUT was developed as a normative theory (a model that suggest how to best make decisions), not as a descriptive theory (a model that describes how people do in fact make decisions). Nevertheless, many economists used EUT to model human choice behavior because they argued that people's behavior approximates EUT's assumptions. Since the 1950s, and more so since the 1970s, researchers have pointed to many cases and scenarios in which people clearly don't behave in accordance with EUT---people seem to violate some of its basic assumptions. Ch 8 describes some of these "violations."
• The bottom line of the Allais paradox is that people interpret probabilities more "qualitatively" than mathematically. The two choices on p. 85 feel very different, even though they are mathematically equivalent. Try to articulate what makes them appear different.
• The bottom line of the Ellsberg paradox is that people don't like options that have partially unknowable consequences (i.e., you don't know how likely the consequences are), even if on average this option would lead to better outcomes than another option that has consequences with a knowable likelihood. Try to find a real-world example that has a structure similar to Ellsberg's paradox---where you choose between one option that has consequences whose likelihood you know and another option that has (attractive) consequences whose likelihoods are not known.
• The key to the unwanted consequences on p.88-89 is the decision rule in the middle of p.88. If you follow that rule of pairwise comparisons, you will violate transitivity. The point here is that people often use those rules without realizing that they lead to suboptimal and unreliable outcomes.
• The section on preference reversals shows again that people don't react to probabilities in a strictly mathematical way. In particular, Slovic (a UO researcher!) and Lichtenstein demonstrate the difference between comparing two options (where people pay attention to each options likelihood) and evaluating each option by itself (where people pay attention to the option's value). This particular study reflects a basic insight of psychology: that people perceive an object in isolation very differently from that same object in comparison to another object.

• In this short paper I step away from the usual perspective of psychology, which looks at phenomena such as judgments and decisions as objective processes/mechanisms. I argue that it is important to understand how people themselves conceptualize judgments and decisions. Once we know more about these concepts as seen from people's own perspective we may more fully understand the phenomena of judgment and decision making.
• Note that decisions can follow premeditation or be spontaneous. This distinction refers to the amount of reasoning that goes on before acting on one's decision. In addition , we can look at how conscious this reasoning process is. ( To use the language of lecture 2, we may anticipate the consequences of our choices with full consciousness or only rather vaguely, without much conscious attention.) When you reason a lot about a decision, it's very likely to be conscious reasoning. If you decide spontaneously, you may be fully conscious of your spontaneous decision or not be very conscious of it. SO even thought I say (on p.7) that there are four types of decisions (premeditated/spontaneous crossed with conscious/unconscious), two type are most frequent: conscious+premeditated decisions and preconscious+spontaneous decisions.
• The relation between the folk-psychological basis of decisions and the famous normative model of expected utility theory should be clear: Probabilities refer to beliefs, preferences refer to desires, and premeditation is the process during which people (supposedly) calculate the expected utility of each option and decide on the one that has the highest utility. It is certainly true that decisions are based on an integration of beliefs (probabilities) and desires (preferences). But we will see that this integration is not well described by a "utility function."

• It is important to keep in mind that expected utility theory was not meant to be an exact description of what people actually do when they make decisions; it should be seen as a normative theory that gives advice how to best make decisions (if certain assumptions are met). In recent decades, psychologists got more and more interested in developing descriptive theories of decision making---theories that do describe and explain people's actual decision making. Chapter 9 reviews some of these descriptive theories.
• We discussed prospect theory in class, and the 1984 article by Kahneman & Tversky also covers the theory.
• Certainty and pseudocertainty were discussed in class as well.
• Regret theory connects back to lecture 2 on dissonance and regret.
• The sections on multi-attribute choice and noncompensatory strategies can be skimmed. But read the sections on "important dimensions" and especially the conclusion.

* Kahneman, D., & Tversky, A. (1984).

• Be sure to distinguish, as we did in class, several features of prospect theory: the gain-loss asymmetry on the value function; framing effects (due to shifts in reference point on the value function); and decision weights.
• Note that the authors give a theoretical explanation of the jacket+calculator example that differs slightly from the explanation we discussed in class. Their has to do with "mental accounting."
• The endowment effect and the difference between losses and costs were also discussed in class. The assigned article by Kahneman, Knetsch, and Thaler (see below) goes into more depths on the endowment effect.

* Kahneman, D., Knetsch, J. L., & Thaler, R. H. (1991).

• Don't worry about the "indifference curve" on p.197. It is not essential for a good understanding of the article. What is most important about this section is the experiment they report on. Be sure you understand the different roles that people were in and how they behaved differently.
• In what sense are choosers and buyers in similar psychological situations? In what sense are choosers and sellers in psychologically different situations?
• Make sure you understand the difference between willingness to pay and willingness to accept. The difference between the two is what the article tries to explain by means of prospect theory (specifically, the notion of loss aversion that follows from prospect theory).
• Follow the argument on p. 197 that shows how "status quo bias" can be seen as a consequence of loss aversion.
• Note that foregone gains are evaluated on the gain part of the value function---you slip from a point of gain back down to the reference point. Actual losses are evaluated on the loss part of the value function---you slip from the reference point down to a point of loss. Because the loss curve is steeper than the gain curve, the slipping of foregone gains is, of course, less steep (and hence less painful) than is the slipping of actual losses.

ch 4: Context dependence

• Plous does a great job in introducing the different context effects. Several more examples of contrast, primacy, and recency effects across many domains were given in class.
• Read the conclusion carefully.

• The distinction between a straight loss and an insurance premium is important. The greater the threat of a large loss (e.g., earthquakes), the easier it is to get people to pay large insurance premiums.
• We will encounter the power of survey demands again later in the course. For now, you can skim the sections on pseudo-opinions and filtering thereof.
• You can also skim the sections on attitudes and on the attitude-behavior relation---they are covered in more detail in Psy 456 (Attitudes and Social Behavior).

• In the section on Open Priorities, Plous talks quite a bit on the influence of response alternative. Relate this topic to the class discussion on standards of comparison and the options available in a particular choice situation (e.g., one CD player or two).
• Framing and psychological accounting are also discussed in Kahneman & Tversky (1984) and in class.

• This chapter is a good overview of social-psychological theories and findings that are pertinent to judgment and decision making. While reading the chapter, ask you yourself which of the phenomena facilitate good decision making and which stand in the way of it.

• This chapter complements the lecture on groups decision processes.

• The chapter gives several examples of how perceptions are influenced by expectations, values, or fears. But it also shows how these "expectation-laden" perceptions are directly transformed into new beliefs, which in turn confirm the initial expectation. This is why chapter 20 (see below) fits so well with this one.

• This chapter describes two related phenomena: Confirmation bias (the tendency to verify rather than falsify hypotheses) and self-fulfilling prophecies (the process by which an initially false belief creates the reality that ultimately makes the belief true). These two phenomena are distinct and must not be confused.
• The Wason task and the Snyder & Swann experiment are examples of the confirmation bias. The "Pygmalion effect" and the Snyder, Tanke, & Berscheid experiment are examples of self-fulfilling prophecies.
• Think about the possibility of self-destroying prophecies. Can you find examples?

• You don't need to know James' or Russell's conception of belief. But you should be able to distinguish the Spinozan and the Cartesian procedure of belief acquisition.
• The "principle of premature output" is important to distinguish between the two models. Why? Compare the Cartesian who gets distracted to the Spinozan who gets distracted. Which one will hold more false beliefs?
• Note the discussion of negative statements (denials); it shows how media might manipulate public opinion without making contested claims.
• How does the Spinozan model explain why people believe so easily what they perceive?
• Think about the socio-cultural influences on a "Spinozan." In a culture that holds people accountable for their claims, the Spinzoan believer will have to spend effort on correcting/rejecting many initially accepted beliefs. In a culture that leaves many beliefs unquestioned, the Spinozan will not spend much effort on those correction processes. In what type of culture do we live? And what subcultures do you know that challenge or don't challenge the Spinozan?

• In a fluent style, Dawes comments on the sunk-cost bias, confirmation bias, and issues of judgment diagnosticity. Collect examples from this chapter, in case you need them.
• Dawes also discusses the only successful technique of convincing someone to give up a cherished hypothesis: offer an alternative hypothesis. One could argue whether the reluctance to give up a belief without having an alternative available is rational or not. At least people are willing to change their mind if you offer a plausible alternative, especially if you provide causal reasons for it.

• The Monty Hall problem is a tricky one. You have only truly understood it if you can explain it to a skeptic (who initially thinks that switching is not a good idea).
• We have discussed the confusion of the inverse extensively in Lecture 8.
• The underestimation of compound probabilities is widespread. Read the example carefully and also relate them to the pregnancy example from Lecture 10.
• The section on perception of risk underscores some points that Paul Slovic makes in a separate article assigned for Lecture 10.

• This chapter introduces you to the major concepts of signal detection theory and discusses the relationships between the "criterion" (beta) on the one hand and the base rates as well as costs/benefits on the other. I derived formula 1.4 (p. 9) in class. At the end of the chapter, the author points to some applications of SDT. In class I mentioned several more (mainly in the domain of person judgment).

• The chapter focuses on people's tendency to see patterns where there are none. Clearly, people don't see randomized series (e.g., coin tosses) as random, but they do see highly alternating series (which are not truly randomized) as "random." Do people perhaps have a different concept of randomness than mathematically oriented researchers?
• In class I argued for the usefulness of having a "lenient" criterion of seeing meaning in apparent randomness. A reaction paper might discuss costs and benefits of such a lenient criterion in different situations.

• This article gives plenty of material to discuss. If you choose to write a reaction paper on it, be sure that you don't misrepresent the authors; be fair to the presented arguments; and account for the collected data.

• The article approaches the topic of risk almost from a folk-psychological perspective by taking seriously how people perceive risks.
• Be sure to distinguish between experts' and lay people's perceptions of risk. What are the major dimensions of the lay concept of risk?
• How could you describe the differences between experts and lay people in terms of (a) prospect theory's decision weights and (b) signal detection theory's beta criterion?

• The major point about this article is that it puts judgment processes into a conversational perspective. That is, it looks at judgments as things that happen in a social context, most often in the course of a conversation. By examining the "logic" (the rules) of conversations, we therefore learn more about the "logic" of judgments.
• In section II. Schwarz discusses the different assumptions people make about their conversation partners (that they are relevant, truthful, etc.) and then, in section III., he applies our understanding of these assumptions to experiments on human judgment. So Schwarz takes a meta-perspective: rather than analyzing human judgment directly, he analyzes experiments on human judgment and looks at the conversational processes within these experiments. And for several of these experiments, the conversational perspective indeed suggests a quite different interpretation than the "heuristics-and-bias" party line would favor.
• Be sure you are able to interpret some of the classic Kahneman & Tversky experiments (e.g., Linda the bank teller, or the engineer/lawyer problem) both from the party line perspective and from the conversational perspective.
• Several of the issues Schwarz brings up as problems of experiments have been discussed in class and in Plous ( ch. 4 and ch 6): effects of response scales, order effects, contrast effects, leading questions, standards of comparison, etc.

• Funder criticizes the heuristics-and-bias approach from several angles. He distinguishes between errors in the lab and mistakes in the real world. He then casts doubt on the generalizability of findings of errors in the lab to mistakes in the real world. (Be sure you understand the difference between errors and mistakes.) And he encourages more naturalistic research into the accuracy (and inaccuracy) of people's judgments.
• In his reinterpretations of classic "error findings" (e.g., overattribution), Funder anticipates some of the criticisms that Schwarz lays out in more detail: the influence of conversational norms on the behavior of research participants.
• In the last section, Funder actually demonstrates research that studies accuracy of social judgment directly and proposes further directions that research can take.

• For the topic of prediction, focus on the sections on "nonregressive prediction" and "clinical versus actuarial prediction," pp. 116-119. The latter section is discussed in more detail in Dawes' article ch 19: Overconfidence
  • The study of confidence and overconfidence compares subjective features of a belief with its accuracy. The subjective feature is how strongly you believe a fact, how convinced you are that you are right. The empiricist philosophers (such as John Locke) argued that false beliefs differ from true beliefs by their subjective feel---true beliefs are more intense. But research on confidence demonstrates that these subjective feel says little about a beliefs accuracy; if anything, we should be particularly careful when people express their beliefs or predictions in extremely strong terms "I am 100% sure..."
  • Note the following difference:
    1. Overconfidence is greater the more confident people are (i.e., at a confidence of 60%, people are typically correct about 50% of the time---10% overconfident---whereas at a confidence of 90%, they are accurate only about 65% of the time---25% overconfident).
    2. Overconfidence is smaller the more accurate people are (i.e., at an accuracy of 50%, people are only slightly overconfident---saying they are 60% confident---whereas at an accuracy of 90%, people may not be overconfident at all---saying they are 90% confident).
  • The discrepancy between people's expressed confidence and their accuracy is partly explained by their notorious difficulty with probabilities. But more importantly, people have a very hard time imagining that they are wrong. In fact, overconfidence decreases entirely when you force people to come up with alternative predictions in addition to their initially preferred one, or when you force them to generate explanations both for and against their preferred prediction (p. 228). This consideration of alternatives allows people to make room for their own fallibility. Here again, we see the benefit of considering alternatives (the "falsifying" approach) , just like we did in Lecture 7 on beliefs.
  On p. 12 of his article, Dawes points out that people prefer an expert's advice to be based on a "burst of intuition" rather than a component summary score. But is this true both for good news and for bad news? Wouldn't you rather hear bad news in the most pallid way possible?
• The last few paragraphs of the article (starting with "Back to Ben Franklin") are particularly important: They link this topic to the larger issue of awareness, self-insight, and the resulting improvements of our own decision making.