Chapter 9
Chapter 9: Thinking
Chapter Review
MENTAL REPRESENTATIONS
- The components of knowledge can be regarded as mental representations. Analogical representations capture some of the actual characteristics of what they represent; symbolic representations bear no such relationship to what they represent.
- Mental images have many picture-like properties, as shown by studies of image scanning. However, other evidence indicates that visual images—like our visual perceptions—are not a simple re-embodiment of some stimulus but are already organized and interpreted in ways that the corresponding picture is not.
- Many of the same brain regions are active during visual perception and visual imagery. Related findings come from work on brain damage: Damage that disrupts vision also seems to disrupt visual imagery and vice versa, adding to the argument that imagery and perception rely on many of the same brain areas.
- Much of our knowledge is represented via propositions, which in turn are likely to be expressed in the mind in terms of associative links among nodes. When you think about a particular topic, activation spreads from the nodes representing this topic to other, related, nodes.
JUDGMENT: DRAWING CONCLUSIONS FROM EXPERIENCE
- In judgment we seek to reach beyond the evidence we’ve encountered, to draw new claims based on this evidence. In many cases, judgment relies on shortcuts called heuristics. The availability heuristic is used for making frequency estimates, relying on the availability of instances as an indicator of frequency in the world. This strategy is often effective but sometimes leads to error, including errors in consequential domains.
- The representativeness heuristic is used for making judgments about categories and amounts to an assumption that the category is homogeneous (so that each instance in the category is representative of the category overall). This strategy, too, is often successful but is overused—as, for example, in “man who” arguments.
- People sometimes rise above the use of heuristics, inviting dualprocess theories. According to many authors, we use System 1 thinking in many situations; but we sometimes use the slower, more effortful, and more accurate System 2. The use of System 2 is more likely if appropriate triggers are in place, or if the person has been trained in how to think about quantitative evidence.
REASONING: DRAWING IMPLICATIONS FROM OUR BELIEFS
- In reasoning, we try to draw implications from our beliefs; this process is crucial for the use of knowledge and provides a means of testing our beliefs. However, people often show a pattern of confirmation bias—taking evidence more seriously if it confirms their beliefs than if it challenges them.
- In showing confirmation bias, people often seem illogical— and consistent with this notion, people often perform badly on simple tests of logical reasoning. In studies involving syllogisms, for example, participants are more likely to judge a conclusion valid if it strikes them as plausible—independent of whether the conclusion follows logically from the stated premises. More generally, the quality of reasoning is influenced by the content of a reasoning task, and performance is much better if logical problems are framed in ways that trigger pragmatic thinking. This is plainly evident in the selection task. People do poorly with the original version of this task, but they reason well with some variants of the task.
DECISION MAKING: CHOOSING AMONG OPTIONS
- In making choices, people are appropriately sensitive to the outcomes of a decision and to the degree of risk associated with each option. However, people are also heavily influenced by how the decision is framed. If the problem is cast in terms of gains, people tend to avoid any risk. If exactly the same problem is cast in terms of possible losses, people seek out risk, presumably in hopes of avoiding the risk.
- People are surprisingly inaccurate in affective forecasting— predicting their own future emotions. People often overestimate how strongly and for how much time they will react to upcoming events. People also overestimate how much regret they will feel if a decision turns out badly.
- People also seem to seek out options, but decision making is often worse when people have too many options. This may be because having options encourages comparisons, and comparisons tend to draw attention to each option’s disadvantages rather than its advantages.
- The factors reviewed here suggest that people often make decisions that do not move them toward the things they value. However, it may be more sensible to evaluate decision making by focusing on the process rather than the outcome, because people often prefer decisions that they can justify or explain. This may not lead people to the best possible outcome, but will allow people to satisfice—i.e., to find satisfactory outcomes.
PROBLEM SOLVING: FINDING A PATH TOWARD A GOAL
- Problem solving is a process that moves us from an initial state to a goal state and depends heavily on how we understand the problem. Often this understanding involves a hierarchical organization—the problem is broken up into subproblems that can be solved via subroutines, some of which are well practiced enough to be automatic.
- Experts rely heavily on subroutines, but they also know more of—and pay attention to—a problem’s higher-order patterns. These attributes are evident in how experts remember problems and in how they categorize them.
- Problem solving is sometimes blocked by an inappropriate mental set. A set can sometimes be overcome by working backward from the goal or by finding an analogy. Sometimes the solution to a problem requires a radical restructuring to overcome a misleading mental set, and restructurings may be an important feature of creative thinking. Some accounts suggest that restructuring often occurs after a period of incubation, although the nature of incubation—if it exists at all—has been a subject of dispute.