Intuition = inference based on a feeling
Evidential intuition = inference based on unconscious assessment of evidence and alternative hypotheses
Non-evidential intuition = inference allegedly based on a sixth sense that does not use or require evidence.
This may involve unconscious processing that largely ignores evidence and instead depends on other considerations such as motivated inference, prejudice, and bogus authority.
Disconfirmation:
Hypothesis H predicts observation O.
But O was not observed.
Therefore, we should consider abandoning H.
But people find ways to hang on to H even when it has been disconfirmed. E.g. religious predictions of the end of the world.
See Gilovich, ch. 3
People have a tendency to base judgments on information that is memorable or vivid rather than reliable.
This can lead to hasty generalization.
Representativeness can lead us to misinterpret evidence, e.g. by assuming that like causes like, and in the conjunction fallacy:
Fred is a UW student who loves to write computer programs and spends much of his time playing computer games. He is awkward around other people, and eats little besides pizza and cola.
Fred is an engineering student.
Fred is an English major.
Fred is an English major and is doing a minor in computer science.
Problem: the conjunction can not be more probable than its conjuncts.
Denying the evidence (Schick, ch. 6)
Tendency to reject evidence rather than to abandon a favored hypothesis with which the evidence conflicts.
Hasty generalization (Schick, ch. 6)
Tendency to make a judgement about a group of things on the basis of evidence concernign only a few members of that group.
Conjunction fallacy (Schick, ch. 6)
Tendency to conclude that a conjunction (A&B) is more probable than one of the conjuncts (A). This occurs when the other conjunct (B) is highly representative or available.
Question: What have your science teachers said is different about science?
Hypotheses derived from observation? Not necessarily: hypotheses can be pure speculation, but must be evaluated with respect to evidence.
Science is falsifiable? Not necessarily: any theory can be held onto if evidence is denied and auxiliary hypotheses are modified.
Science is testable, i.e. predicts something new? Not necessarily: some theories like the early theory of evolution are qualititative and do not make sharp predictions.
(from P. Thagard, Computational Philosophy of Science, 1988).
| Science | Pseudoscience |
| Uses correlation thinking | Uses resemblance thinking (representativeness) |
| Seeks empirical confirmations and disconfirmations | Neglects empirical matters |
| Evaluates theories in relation to alternative theories. | Oblivious to alternative explanations. |
| Uses broadly explanatory and simple theories | Non simple theories with many ad hoc hypotheses |
| Progresses over time: develops new theories that explain new facts | Stagnant in doctrine and applications |
Testability: predicts something new
Fruitfulness: open new lines of research
Scope: how much the theory explains
Simplicity: make fewest assumptions
Conservatism: fits with established beliefs
In sum, a theory is accepted only if it is the best explanation of all the available evidence.
Pseudoscience = field that claims to be a science but is not.
Example: astrology vs. astronomy and psychology
Example: creationism vs. evolutionary biology
Example: parapsychology (ESP) vs. psychology
Alternative explanations for ESP results: fraud, sloppiness, chance.
Pseudoscience links:
Updated Feb. 10, 2003