3: Introduction to Scientific Modeling

Question 1

scientific model

Answer 1

make inferences about a target of interest

always involve idealisations (unrealistic assumptions about the target which are departures from reality)

Question 2

relationship between scientific theories and models

Answer 2

“models as mediators”

- theory cannot be used directly to make inferences about systems

Question 3

what is data really?

Answer 3

data is theory-ladden
- whatever data you have is affected by theoretical presuppositions

data is never theory-free and implicitly involves a lot of theoretical assumptions

Question 4

sensitivity of a test

Answer 4

probability that if someone has antibodies, the test will come out as positive

the higher it is, the more accurate the test is

Question 5

specificity of a test

Answer 5

probability that if someone doesn’t have antibodies, the test will come out negative

the higher it is, the more accurate the test is

Question 6

Quine-Duhem thesis

Answer 6

theory/model can never really be falsified

• falsifiability is not what makes a theory/model scientific
• if things can only be scientific when falsified, nothing is scientific

what you are testing is not just the model but the model and auxiliary hypotheses (hypotheses about how the data relates to the phenomena of interest)
- instead of rejecting the model, you reject the hypotheses and therefore the data