Task 4 - Reasoning Flashcards
Inductive Reasoning
- Forming generalisations (which may be probable but not certain) from examples (specific –> general)
- Conclusion can contain new information
- -> Inductive strength – argument has strength if it is improbable (not impossible) for the premises to be true & the conclusion to be false
- Probable truth
Analogical reasoning
in which an individual tries to solve a current problem by retrieving information about a similar problem that was successfully solved in the past
Hypothesis Testing
2-4-6 Task (Wason)
- people are given a bunch of numbers and they have to determine the rule only by offering examples, not by asking direct questions about the rule
- confirmation bias – (fallacy of affirming the the consequent)
Deductive Reasoning
general –> specific
- Conclusion represents info that was already implicit in the premises (no new info added)
- Allows us to draw conclusions that are definitely valid, assuming that premises are true
Deductive validity
argument valid only if it is impossible for the premises to be true and the conclusion to be false
– When premises are true and you reason according to logical principles, your conclusion cannot be wrong
Types of Deductive Reasoning
- Propositional
2. Syllogitic
Proposition
claim that can be either true or false
compound propositions
more complicated propositions formed by using logical connectives
Modus ponen
if A, then B and A is given then we can validly infer B
Modus tollens
if A then B and we know B is false we know A is as well
Denying the antecedent
fallacy in which the info that the antecedent isn’t true leads to the conclusion that the consequence isn’t either
Affirming the consequent
fallacy in which the consequence is thought to be true because the antecedent is said to be true
Wason’s selection task
must affirm modus ponies and pollens
Categorical syllogism
3 premises that deal with classes of entities (include quantifiers)
premises
propositions about which arguments are made
belief bias
tendency to accept invalid conclusions that are believable rather than believe valid conclusions that are unbelievable
Effects of premise phrasing
– Patterns of Reasoning Performance
- the way information is stated can make a reasoning task easier or harder
- More errors when premises have negatives in them
- Quantifiers such as “all” or “none” are easier to deal with than “some”
- Syntactically complex statements require more processing resources that are then missing for other reasoning processes
Alteration of premise meaning
– Patterns of Reasoning Performance
- people often misinterpret premises & fail to think of all possible meanings
- Everyday understanding of words such as some, all & if-then are so powerful that people have difficulties ignoring the fact that in reasoning tasks they’re defined slightly different
Content effect
– Patterns of Reasoning Performance
- people reasoning with exactly the same kind of premises will perform differently depending on what the premises are about (e.g. four-card task/ Wason’s selection task with drinking vs. age instead of numbers vs. letters)
- Memory-cueing explanation – own experience lets one think of combinations
Believability effect
– Patterns of Reasoning Performance
- people are likely to judge as valid any conclusion that reinforces their initial assumption, regardless of whether the conclusion follows from the premises
Approaches to the study of reasoning
- Componential approach (Sternberg)
- Rules / heuristics approach
- Mental models approach (Johnson-Laird)
- Heuristic-Analytic Theory (Dual Systems) (Evans)
Componential approach (Sternberg)
- Study reasoning by analysing a task into its component cognitive processes (mental subroutines
- Reasoning as just another mental activity like problem solving & decision making
- Performance components – individual cognitive processes (encode –> recognise infer relationship –> apply relationship
- Meta-components – executive processes used in planning & monitoring (e.g. select which performance components will be used)
- Knowledge acquisition components – used whenever we acquire new information (e.g. elective encoding, combination & comparison)
Rules / heuristics approach
& Cheng vs Braine vs Cosmides
treat reasoning as a special mental process –> rely on special-purpose mental rules (like grammar) for inferences
- —> Cheng: people interpret problems in terms of what they are about and based on that use different schemata
- —> Braine: people can & do use the same set of abstract rules in all situations
- —> Cosmides: Social contract theory
Social contract theory
evolutionary adaptive mechanism for reasoning
- especially good at tasks when content is construed in terms of social costs & benefits
- people have rules maximizing ability to achieve goals in social situations
Deontic rules: easier to understand (rule more explicit)
Mental models approach (Johnson-Laird)
Use deductive reasoning that may be affected by real-world knowledge
- -> Deductive reasoning problems requiring the construction of several mental models are harder to solve because of the increasing demands on working memory
- Principle of truth – people minimise the load on working memory by tending to construct mental models that represent explicitly only what is true & not what is false
- —- Common source of error: failure to construct relevant models, failure to assess the implications of all the models found, failure to construct enough models
Limitations
Mental models approach (Johnson-Laird)
- Ignores individual differences & dual systems
- Doesn’t explain which pieces of background knowledge are used
Heuristic-Analytic Theory (Dual Systems) (Evans)
- Initially use real-world knowledge that may be affected by deductive reasoning through system 2
The theory assumes that reasoning and judgment are facilitated by the formation of epistemic mental models that are generated one at a time (singularity principle) by preconscious heuristic processes that contextualize problems in such a way as to maximize relevance to current goals (relevance principle). Analytic processes evaluate these models but tend to accept them unless there is good reason to reject them (satisficing principle)
- System 1: heuristic processes (implicit)
- -> Makes use of task features, goals & background knowledge to construct a single hypothetical possibility/ mental model
- System 2: analytic processes (explicit)(time consuming, effortful)
- -> May intervene to revise/ replace model
- -> Likely when instructed to use abstract/ logical thinking, participants are highly intelligent, sufficient time available
Principles
Heuristic-Analytic Theory (Dual Systems) (Evans)
- Singularity principle: only a single mental is considered at any given time
- Relevance principle: the most relevant mental model based on prior knowledge & current context is considered
- Satisfying principle: current model is evaluated by system 2 and accepted if adequate (often leads people to accept conclusions that could be true but aren’t necessarily)
Limitations
Heuristic-Analytic Theory (Dual Systems) (Evans)
- Oversimplification to make a distinction between the systems as in implicit vs. explicit –> both can be both
- Says they are various analytical processes but doesn’t elaborate
Brain
Prefrontal cortex (esp. left) -- plays a role in a person's ability to integrate relations (to build mental representations that that incorporate multiple propositions / relationships) - left inferior frontal gyrus (Broca's area!) --> deductive reasoning left dorsolateral prefrontal gyrus --> inductive reasoning
natural explanations
appeal to “observable & empirically verifiable phenomena of the physical/ material world”
supernatural explanations
appeal to phenomena that “operate outside of/ are distinct from the natural world”
supernatural explanations
appeal to phenomena that “operate outside of/ are distinct from the natural world”
Explanatory coexistence
when natural & supernatural explanations are generated to interpret the same events
- Synthetic thinking: loose integration of natural & supernatural frameworks without in-depth consideration of how they might interact
- Target-depending thinking: the 2 explanations are used to account for distinct aspects of a given phenomenon & involves different kinds of causality
- Integrated thinking: explanations become well-coordinated, use both explanations to interpret multiple levels of causality
Human origins
Results of study
no participant provided evidence of explanatory coexistence
Death
Results of study
mostly natural processes as reasons for death and what happens to the body after
- Some target-dependent thinking – supernatural explanation for what happens after death (spirit & soul life) seen as distinct process from cause of death
- Causes of death are much more observable thus more natural explanations
Illness
Results of study
explanatory coexistence in 77% of the cases
- Synthetic thinking: generally natural explanations but also said people can get sick from their behaviour & the way they treat others but couldn’t explain how
- Given greater access to info about proximate causes for illness + the influence of Christianity
Discussion
of the study
Death & Illness: people are often unsatisfied with natural explanations of death & illness and seek to understand why these things happen to them or their loved ones which explains why death & illness show more explanatory coexistence
Base rate effect
Biased through the probability that following statements could be false or true
