Thinking and reasoning

Definitions and Key Questions

• Measuring, stirring and baking etc. are the processes involved in making a cake, but we also need ingredients (e.g. eggs, flour, sugar)

• Decision-making, judgements, reasoning and problem-solving are processes involved in thinking.

Mental representations

• information used for thinking

• stored in LTM

Thinking

• manipulation of mental representations

Concepts & categorisation

What are concepts

• structures in semantic long-term memory

• repositiories of knowledge about categories

Categorisation in the process- concepts are the product

Examples

• an apple is a fruit

• a tiger is a mammal

Use concepts for reasoning, language, understanding

Why are concepts important?

• cognitively effective way of representing our knowledge of the world and its objects

• allow us to make accurate predictions

• categorisation aids communication

• real world uses (legal issues, medical diagnoses)

“To categorize is to render discriminably different things equivalent, to group objects and events and people around us into classes, and to respond to them in terms of their class membership rather than their uniqueness” Bruner et al (1956)

Theories of categorisation

Classical approach, prototype approach, exemplar approach

Classical approach

Items belong to certain categories because they have certain properties in common

Necessary properties - even category member possesses it

• being an unmarried man is necessary for being a bachelor

Sufficient properties - anything that possesses those properties must be a member of the categoru

• any unmarried man is a bachelor

Typically gradient

• some members considered more typical of a category than others

• faster judgements for typical categroy members

Unable define necessary features of some concepts

• e.g. game

Assuments individual features are independent of each other

• features are often correlated with each other

Prototype approach

Rosch and Mervis

• categories have a central description

• a prototype is a ‘best’ category member

Family resemblance

• no single attribute is shared by all examples

• each example has at least one attribute with another category member

Exemplar approach

Many instances of a category are stored in memory (exemplars

• no pre-stored prototype

• compare item with all stored exemplars

Can account for the variability of instances within categories and inability to list necessary features, but has a heavy memory load

Problem-Solving

Activity where need identify solution to a problem

• can be easy or more complex

• occurs regularly in real-life

Approaches depend on

• the problem environment

• the problem solver

Involves all aspects of cognitive processing

• memory

• attention

• perception

• other aspects of thinking

What is a problem?

Start state → set of available actions → goal state

Methods of investigation

Outcome measures:

• Success/Failure?

• How many moves?

• How long did it take?

• How effective?

Process measures:

• Look at strategies employed during problem-solving

• Self-report inventories

• Protocol analysis – verbal account given as solution is carried out

Simple problem-solving

Knowledge-lean: do not require extensive background knowledge

Thorndike (1898)

• trial-and-error learning

Gestalt approach

• reproductive vs productive thinking

• concept of insight

• kohler (1925): ‘sultan’ and his use of tools (insight)

The Candle Problem - Functional Fixedness

Duncker (1945)

You are given a box. It contains:

• Candle

• Tacks

• Matches

Your task:

• Attach the candle to wall

• Light it so doesn’t drip onto table below

The Water-Jars Problem - Mental Set

Luchins and Luchins (1959)

• How measure out 5 litres exactly, how measure out 25 litres exactly.

Gestalt Approach

• Subjects become fixed on function of objects from prior experience

• Result failure to think creatively beyond object’s perceived function (gestalt).

• This effect is also found when become set in our ways.

• Continuously re-apply a way of solving a problem

• Even when there is more parsimonious way of achieving goal (e.g. Luchins & Luchins, 1959)

Summary:

• Problems need to be restructured and also can involve insight

• Showed mere reproduction of learning insufficient

Complex Problem-Solving

Require extensive knowledge base

Difference between a noice and expert:

• Knowledge base

• Problem representation

• Strategies

• Problem-solving schemas

Research used domains such as chess to explore differences between experts and novices

Chess Expertise - DeGroot (1965) and Chase & Simon (1973)

Pattern Recognition, Efficient Search, Domain memories

• Expert vs. novice chess players

• Asked think aloud as studied board & chose moves (protocol analysis

• Asked remember placements of pieces on chess board & then reconstruct it (recall-reconstruction task)

Expertise and Practise

• 10 years for expertise in violin, chess, maths etc. (Ericsson et al, 1993)

• 3,000 hours to be expert and 30,000 to be chess master (Simon & Chase, 1973)

Practice improves performance because:

• Can work in greater chunks

• Automaticity

• Restructuring

• Not only practice – role talent

Reasoning + Judgement

Decision-making

• selecting one out of a numver of presented options or possibilities

Judgement

• component of decision-making that involved calculating the likelihood of various possible events

Reasoning

• forming conlcusions, inferences or judgements from given information

Can use formal (deductive) reasoning techniques

• based on principle of logic

• involved premises and conclusions

Tversky + Kahneman (1972)

• Many studies show biases in reasoning & judgement

• Suggests processes not rational

• Limited mental processing capacities

• Use strategies of simplification to reduce complexity of reasoning & judgements

Mental heuristics

• Mental ‘rules of thumb’ used to simplify reasoning

• representativeness, anchoring, availability

Representativeness:

involves estimating the likelihood of an event by comparing it to an existing prototype that already exists in our minds

• this prototype is what we think is most relevant or typical example of a particular event or object

Anchoring:

How many African nations (%age) are there in the U.N.?

Wheel 10 → est 25%

Wheel 65 → est 45%

Availability:

Consider the letter K. Is K more likely to appear as the first letter in a word or as the third letter? Tversky & Kahneman (1973)

Lichtenstein et al (1978)

• Relative likelihood of different

causes of death

• Murder rated more likely

than suicide

Lawyer, doctor, politicians (affairs)

Numerosity:

perceptual phenomenon relating to numerical cognition

• describes the tendency for humans to be drawn to bigger numbers even if they apply to small things

Which pile of coins has higher monetary value?

Evaluation

• Reasonable evidence for many heuristics proposed by Kahneman & Tversky

• However…

• Theories do not specify when and how heuristics are used

• Work focuses on errors as a result of using mental heuristics

• Lack of ecological validity

Fast and Frugal Heuristics

Todd & Gigerenzer (2007)

• Heuristics are valuable & can speed reasoning/judgements

• Doesn’t always make it more accurate, just quicker

Example:

• Which of Herne and Cologne has larger population?

• Take the best strategy/heuristic

• Search rule – Stop rule – Decision rule

• Recognition heuristic is one of most commonly used

Decision-Making

Two Systems of thinking:

System 1: Intuitive, automatic, immediate, effortless, often emotionally charged, difficult control (Kahneman, 2003, 2011)

System 2: analytical, controlled, rule-governed, slower, serial, effortful

Theories of Decision Making

Normative theories

• Ideal/rational decision-making

• People do what they ought/should in a given situation

Descriptive theories

• Characterise how people actually make decisions

Prescriptive approach

• Straddles normative & descriptive theories

• Investigates how to help people make better decisions (e.g decision analysis)

Subjective Expected Utility Theory

Savage (1954)

• Normative model of risky choice

• Rational decision-maker trades off value of all possible outcomes by the likelihood of obtaining them.

SEU = P (outcome) x Utility

Prospect Theory

Kahneman and Tversky (1979)

• People evaluate decision outcomes in terms of gains and losses

• Act as to maximise gains and minimise losses?

• More sensitive to losses than gains (loss aversion)

• Differences between people with low and high self esteem (Josephs et al, 1992)

Social Functionalist Approach

• SEU and prospect theory generally focus on behaviour in laboratory experiments

Tetlock (1991)

• “Subjects in laboratory studies… rarely feel accountable to others for the positions they take. They function in a social vacuum… in which they do not need to worry about the interpersonal consequences of their conduct” (p.453)

Sunk-cost effect (Simonson & Staw, 1992)

• People are likely to put more money into something that has already proven fruitless

• Attempt to justify previous investment

• More likely when justification to others required

Decision-Making and Neuropsychology

Iowa Gambling Task (Bechara et al, 1994, 2000)

• Neurotypical individuals = rational performance

• Damage to ventromedial prefrontal cortex → worse at task

Summary

What are mental representations? How are they involved in thinking? What form do mental representations take?

What are concepts? How do we categorise objects into concepts?

Problem-solving? Simple- and complex-problems

Are humans rational when making judgements & decisions? Do they always use formal reasoning & probabilities to inform these processes?

If not, then what processes are involved when we make judgements & decisions?