PSYC 2022 Lecture Notes - Wason Selection Task, Inductive Reasoning, Deductive Reasoning
18 views2 pages
Lecture 6 PSYA02 Notes
•Inferring specific instances from general principles
•An argument describing the relations between categories of
–Premise 1: All A are B All VWs are reliable.
–Premise 2: C is an A The Beetle is a VW.
–Conclusion: C is B The Beetle is reliable.
•An argument describing the conditional relations between events
–Premise 1: If P then Q If it is a VW, then it is
–Premise 2: P is true The Beetle is a VW.
–Conclusion: Q is true The Beetle is reliable.
Errors in Deductive Reasoning
•Social Contract Theory (Tooby & Cosmides)
oEvolutionary selection pressures have equipped us with a “cheater
detection mechanism”—an innate set of inferential rules that help us to
detect the violation of social contracts
oEvidence: Wason Selection Task performance improves when the task is
framed in terms of contract violation
•Inferring general principles from specific instances
oGeneral Induction: known instances → *all* instances
oSpecific Induction: some instances → other instances
oHypothesis—a proposition that can be evaluated or tested by gathering
evidence to support or refute it
oNo inductive process can ever be certain: we cannot know all the
instances that may exist, any one of which may disprove the
Errors in Inductive Reasoning
oThe disinclination to seek evidence that would indicate whether a
hypothesis is false
oE.g., the Wason 2–4–6 Task
•Participant’s goal: Discover the rule
•“Even numbers increasing by two:” e.g., 8–10–12
•“Any set of numbers increasing by two:” e.g., 7–9–11
•Participants then switch from confirm to disconfirm, and soon
discover the surprisingly simple rule: “Numbers of increasing
•A special kind of inductive reasoning; the process of applying knowledge from
domain (“the source”) to another domain (“the target”).
•“Can the solution for one problem be used to solve (i.e., be mapped
Problem -- Biological Virus Computer Virus
Solution -- Vaccine ?
oHold target in STM; access source from LTM
oMap features of the source onto those of the target
oDecide whether or not the analogy is likely to be useful
oIsolate the structural features shared by the source and the target
oHypothesize about the target from what is known about the source
Errors in Analogical Reasoning
•Source: General & Army
oEvil dictator mines all roads to fortress
oAttacking army will be blown up if too large
oAttackers break up into small groups, overcome dictator
•Target: The Radiation Problem
oPatient has an inoperable stomach tumor
oDoctors have a laser that can burn the tumor out
oLaser will also burn healthy tissue: What should you do?
•Participants Don't Recognize Relationship
oW/O hint: ≈ 20% use convergence (Gick & Holyoak,1980)
oWith hint: ≈ 90% use convergence (Holyoak & Thagard, 1995)
•Deductive reasoning allows us to draw inferences from general principles to
specific instances. Deductions are limited by whether the premises of our
arguments are true.
•Inductive reasoning allows us to draw inferences from specific instances to
general principles. Inductions are limited by the possibility of disconfirming
•Analogical reasoning allows us to apply knowledge from one domain to
another. Analogies are limited by our inability to see beyond surface
Inferring specific instances from general principles: categorical syllogism, an argument describing the relations between categories of things. The beetle is reliable: conditional syllogism, an argument describing the conditional relations between events. Premise 1: if p then q reliable. If it is a vw, then it is. Inferring general principles from specific instances: general induction: known instances , specific induction: some instances , hypothesis a proposition that can be evaluated or tested by gathering, no inductive process can ever be certain: we cannot know all the. *all* instances other instances evidence to support or refute it instances that may exist, any one of which may disprove the generalization. Errors in inductive reasoning: confirmation bias, the disinclination to seek evidence that would indicate whether a, e. g. , the wason 2 4 6 task hypothesis is false. Any set of numbers increasing by two: e. g. , 7 9 11.