Categories & Concepts

• Category: Collection of objects, events, or ideas.
• Concept: The mental representation of the category; the information in the mind that allows us to group together the things in a category.

Functions of Concepts

• Classification
• Understanding & explanation
• Prediction
• Reasoning
• Communication

Defining Features Model

• Meaning of a concept is given by features or properties.
• Features are necessary and sufficient.
• Works for mathematical and scientific concepts but not for many other concepts: What are the defining features of "chair?"

Prototype Model

• Meaning of a concept is given by features or properties.
• Membership in a category is determined by the number of shared features.
• "Good" members share many features with other members and few features with nonmembers.
• Typical features form a prototype, or "ideal" member of the category.

Example: Diagnostic and Statistical Manual of Mental Disorders (DSM)

• A person can be classified as depressed if he or she is extremely unhappy and has any 5 of 9 symptoms (i.e., features) present every day for 2 weeks.
• Hence, two people with the same diagnosis of "depression" may share only 1 of 9 symptoms.

Exemplar Model

• Meaning is the set of examples of the category, which are called "exemplars."
• Membership is determined by similarity.

Which model is correct?

• People have very good knowledge about specific examples of categories; exemplars must be stored in memory.
• But we also have the impression of having abstract knowledge about certain concepts.
• Both models are probably true.

Problem Solving

• Behavior directed toward achieving a goal.
• It involves breaking the goal into subgoals.
• The means of achieving subgoals must be selected.
• Example: Köhler & Sultan the chimp.
• http://www.ling.lancs.ac.uk/monkey/ihe/linguistics/LECTURE4/4prob.htm

Methods of Problem Solving

• Means-ends analysis
• Problem solving by analogy

Means-Ends Analysis

• Actions are taken to reduce the gap between the current state and the goal.
• New actions may themselves become subgoals.
• Example: Tower of Hanoi

Problem Solving by Analogy

• The structure of a problem whose solution is known is used to solve a problem whose solution is unknown.
• Example: Duncker's radiation problem.

Gick and Holyoak's Experiment

• No analogy or hint 8%
• With analogy but no hint 20%
• With analogy and hint 92%

"Inert Knowledge"

• People often have the knowledge they need to solve problems, but don't know that they have it or don't know how to apply it.
• A major problem in education.

How can analogical problem solving be improved?

• Increasing the similarity between problems
• Experience

Problems in Problem Solving

• Inflexible representations (functional fixedness)
• Hidden representations (the "monk problem")
• Set effects (Luchins's water-jug problem)
• Cultural blocks

Decision Making

• Framing
• Representativeness
• Availability

Representativeness Heuristic

• Outcomes or conclusions are judged to be likely if they share features with the evidence or premises.
• Examples:
• B B B B B B vs. B G G B G B
• The gambler's fallacy and "law of averages"

Availability Heuristic

• An outcome or conclusion is judged to be likely if examples are easy to retrieve from memory.
• Example: Evaluations of graduate schools.

Summary

• Concepts are the building blocks of human thought, and seem to consist of abstract and specific knowledge.
• Problem solving is goal directed and involves breaking the goal into subgoals.
• Problem solving and decision making rely heavily on heuristics, and can be error prone.
• Heuristics, however, are quick and economical; the benefits must outweigh the costs.