Statistics Theory L1 = Philosophy Of Science Flashcards
Philosophy of science attributes? (3)
Involves:
- Why scientists do what they do.
- Why they use the methods of inquiry they do.
- Question assumptions that scientists take for granted.
What’s the typical problem in philosophy of science?
It’s that there are different perspectives of how science should be conducted.
What are the perspectives on how science should be conducted? (2)
- Logical positivists.
- Scientific paradigms.
What’s the point of these perspectives?
To show you that science is complicated and is not as objective/clear-cut as you were taught.
Proponent of Logical positivists?
Karl Popper.
Proponent of Scientific paradigms?
Thomas Kuhn.
Logical positivists?
= argue that because science is objective, rigorous & based on measurements, it is an ideal tool for seeking the truth about nature.
Logical positivists attributes? (3)
- Explanation should be falsifiable.
- Explanation should be posed in a way that leads to a prediction, which can be compared with evidence, and demonstrated to be false.
- If a theory doesn’t satisfy this condition, it would be called pseudoscience.
Egs of Logical positivists? (3)
- Freudian psychoanalysis (pseudoscience).
- Skinner (real science).
- Einstein (real science, testable).
Pseudoscience?
= theory/explanation changes according to the evidence.
Real science?
= theory/explanation remains constant despite changes in the predictions.
What if observations don’t match predictions? How much evidence is necessary to reject a theory?
Consider the following:
- Evolution vs Intelligent design.
- Darwinism vs Lamarkism.
Scientific paradigms?
= argue that science is driven by scientific paradigms.
Paradigm?
= a set of fundamental theoretical assumptions accepted by the scientific community.
Paradigm is also?
= a set of exemplars solved by those assumptions.
Paradigm attributes? (2)
- Involves indoctrination.
- Drive the evidence that support it (the indoctrination), thus competing paradigms can’t be compared.
What is scientific reasoning?
= it’s how scientists make conclusions about the processes that shape the world.
How do scientists make conclusions about the processes that shape the world?
They make inferences, i.e., propose processes that can’t be observed through patterns that can be observed.
How do we get reliably from observable patterns to unobservable processes?
Through:
- Induction (mostly).
- Deduction (mostly).
- Hypothetico-deduction.
- Retroduction (not main focus).
Induction?
= when one uses the observed pattern (s) to determine the unobserved process causing those patterns (specific to general).
Deduction?
= when one uses the process to determine the possible patterns that will be produced.
Hypothetico-deduction?
= tests the proposed mechanisms.
Retroduction?
= explains how/why a process produces a pattern (i.e., via a mechanism).
Egs of induction vs deduction? (2)
- Sherlock Holmes.
- Smoking causes cancer.
Which form of reasoning is stronger or more reliable?
In reality we use both.
Forms of scientific reasoning? (4)
- Induction.
- Deduction.
- Hypothetico-deduction.
- Retroduction.
Steps to scientific reasoning? (5)
1) Pattern of interest.
2) Form a question of the pattern.
3) Potential explanation.
4) Form predictions.
5) Test predictions.
Steps 1-3 of scientific reasoning represent?
- Induction.
- Retroduction.
Steps 3-4 of scientific reasoning represent?
Deduction.
Steps 4-5 of scientific reasoning represent?
Hypothetico-deduction.
Using the Oscar Pistorius trial:
Step 1?
Gun, bullet holes, blood, dead girl, screams/cries, confession.
Using the Oscar Pistorius trial:
Step 2?
Why did he do it?
Using the Oscar Pistorius trial:
Step 3? (2)
1) Mistaken intruder.
2) Deliberated, premeditated.
Using the Oscar Pistorius trial:
Steps 4-5? (4)
- More investigation.
- Go back to the crime scene.
- Go out to question witnesses.
- Run experiments in the crime lab.
Correlation?
= not enough to conclude with certainty.
Why do scientific reasoning this way (why this method)? (2)
- We don’t see the true process.
- To check biases.
