Chapter 5: Evidence Flashcards
Evidence
any new fact that supports a possibility under investigation (possibility is hypothesis = H)
Evidence Test
is this fact more likely if H is true or if H is not true?
Independent
suppose that one claim is true. Does that affect how likely the other claim is? If not, the two claims are independent.
The First Rule of Evidence
evidence should increase our confidence in a hypothesis
The Strength of Evidence
A test that involves comparing the following:
A) suppose H is true: how likely is this?
B) suppose H is false: how likely is this?
Comparing these two is the strength test
Strength Test Extremes
Evidence is maximally strong: the evidence entails the hypothesis, so the inference is deductive
Evidence and Probability
Likelihood of claim: degree of confidence
Confidence can be expressed on spectrum ranging from 0 to 1
0 = certainty that claim is not true
1 = certainty that claim is true
0.5 = equal confidence in claims being true or false
What is NOT Meant by Probability
Probability is not restricted to future events
Probability is not a statistical generalization about classes of events
Probability of a claim is simply your degree of confidence in that claim
Degree of Confidence
Degree of confidence is expressed as a percentage, fraction, or decimal between 0 and 1
Probability of H and probability of not-H should add up to 100%
Importance of Strength Test
To test suppositional strength
Forces us to consider the opposite (what would we expect to observe if H were false?)
Prevents one-sided testing
One-Sided Strength Testing
Only paying attention to how likely the evidence would be if the claim were true than if it were false
Opposite Evidence Rule
If E is evidence for H, then not-E is evidence for not-H
Selection Effects
factors that systematically select which things we can observe (makes evidence unreliable)
Confidence in a Hypothesis Depends on Two Things:
- How strong the evidence is
- How confident we were before getting the evidence
Survival Bias
an error in which attention is paid only to those entities that have passed through (or “survived”) a selective filter.