Reasoning Part 2 (Pre lecture 17) Flashcards
Wason selection task (WST)
Peter Wason (1966).
Used to study performance on logical reasoning for conditional propositions/arguments/problems.
Comprises 4 playing cards:
- two cards each showing vowel or consonant.
- two cards each showing an odd or even number.
Conditional proposition (if P then Q): if a card has a vowel on one side, then it has an even number on the other.
WST original
eg. E K 4 7
ppts asked: which two cards must be turned over to discover whether following rule is true?
Choice of cards determines performance.
Performance on original
Typical results: approx 70% choose E and 4 cards- not correct.
Choosing E card is logical because it permits affirmation of antecedent conditional inference.
But choosing 4 card only confirms it. You want to falsify it.
If rule is true you should turn over 7 and expect to see a consonant.
Logically correct E and 7- can check whether true or false. (P and not Q).
Task performance
Why are logically-valid cards (E and 7) not chosen often?
- tend to choose cards that confirm instead of checking if true or false.
Built in “confirmation bias”- tendency to support the rule given.
Psychological evidence (eg. self esteem) to support this confirmation bias.
Poor performance due to its abstract content and lack of context (Manktelow, 1999/2012).
- when you change the task to a more real world context, logical reasoning performance improves. (eg. drinking age version)
Drinking age version
Rule: if person is drinking beer (p), they must be over 18 (Q).
Beer Coke 16 22
Logically correct choices = Beer and 16.
Results: approx 90% chose correct cards most of the time.
Why improve? - people are familiar with real world rules and laws, so it is not difficult to test this hypothesis.
Negation = contradiction or denial of something
Research evidence of reasoning being poorer when propositions involve negation (‘not’).
Wason (1965) used sentence verification task- ppts had to draw conclusions about whether the positively or negatively worded sentences were correct or not.
- found ppts were easier with positive arguments; slower at drawing correct (logical) or incorrect (non-logical) conclusions for arguments involving negative statements (like ‘not’).
Hard to reason about negation because need to consider alternative- takes time and cognitive resources; may draw wrong conclusions.
Issues
Competence: people with no training don’t do well. Just above chance level.
Bias: respond to non-logical features.
Content: eg. drinking age.
Knowledge: prior experience affects.
What is rationality?
Being able to reason correctly/logically.
Not what conclusions we draw (accuracy) but how we draw these conclusions.
To do with methods of thinking.
Bounded rationality
Theory to explain that we are rational beings because of cognitive constraints. Some external factors also- lack of time, lack of motivation.
Two constraints really studied- limited attention span, limited working memory capacity.
Simon (1957)- people don’t do too bad in reasoning in terms of every day life, even if it violates normative standards but could achieve personal goals.
- rationality within confines/boundaries we do well in.
Reasoning is rational to some extent.
Don’t have unlimited cognitive resources to understand everything.
Non-cognitive factors arguably at play- eg. motivation.
Two types of rational reasoning?
eg. Evans & Over (1996); Slowman (1996): proposed idea.
System 1: reasoning according to achieving one’s goals and meeting one’s needs (seen as non-logical/non-rational).
- goal directed rationality.
- rapid, automatic, pre-conscious.
- aware of outcome of thought process but not process itself- influenced by beliefs, experience etc.
- doesn’t conform with principles of logic.
System 2: more slow, laborious, logical. Involved in planning. Reasoning according to normative standards.
- more traditional, rule based rationality.
- eg. task planning, hypothetical thinking.
- aware of thought process as well as outcome.
Distinct systems, but can be used together.
Evidence (Evans & Over, 1996)
Belief bias effect: talking about categories. If ppts believe in syllogism (content of the problem), their belief will override logical reasoning ability.
Relying on ‘gut reasoning’ consistent with system 2 not 1.
Explicit and tacit thought process?
- tacit: task specific and resistant to training. (eg. musical instrument- once knowledge gets to tacit stage you can’t necessarily report what’s going on).
- explicit: general and depend on practice/training (eg. driving a car- you can talk through what you’re doing).
Abstract version of WST: people take longer to inspect cards they choose, and often chosen before they’ve understood the rule.
Non-conscious process for Wason task card choice, then conscious logical justification process.
Prior believability of claim’s conclusions affects logical reasoning performance- belief bias.
Interpretation
Not following study instructions evidence of non-rationality in lab but not elsewhere (Evans & Over)- will poor performance transfer to real life?
System 1: aware of outcome but not process (Sloman, 1996).
Possible ways to test dual systems:
- time pressure to induce 1.
- instructions to induce 2.
Dual systems theories
Provide neat explanation of how we reason in everyday life.
Explains why reasoning often non-rational when judged against principles of logic.
But little new empirical support.
Are we non-rational?
Depends how you define it.
Typically don’t explicitly conform to logic principles in daily reasoning.
We often draw correct conclusions and make good decisions about complex things.
Generally rational by systems 1 standards.
