Phil 145, Week 5

5a Evidence and inference

Intuition, continued

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.

Denying the evidence


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.

Confirmation bias

See Gilovich, ch. 3

Availability error

People have a tendency to base judgments on information that is memorable or vivid rather than reliable.

This can lead to hasty generalization.

Misinterpreting evidence

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.

Error tendencies

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.

5b Science and Pseudoscience

Question: What have your science teachers said is different about science?

How is Science Different from Nonscience?

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.

Profiles of Science and Pseudoscience

(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

Desirable properties of theories

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

Back to Phil 145