Lecture 1
PoS: Investigating the social sciences
Social sciences: Investigating the social world
Demarcation problem: is about drawing a clear line between what counts as science and
what does not. > De grens is vaag tussen wat wel science en wat niet is
What makes something scientific? (demarcation problem)
And you have two main answers:
● Logical positivism → verification > Empirically verified
● Popper → falsification > Can we prove it wrong?
Logical positivism: A statement is only meaningful if it can be empirically verified.
- “Water boils at 100°C” → scientific
- “God exists” → not scientific (not testable)
Key elements Logical Positivism
1. Verifiability criterion
● Meaning = method of verification
● If you can’t test it → it’s meaningless
● To understand a statement = knowing how you would check if it’s true
2. Ideal language of science
● A statement only has meaning if you can test it in the real world.
● Use logic + math
● Make science precise and objective
● What gives a statement meaning in positivism?”
● Correct answer:
→ its method of verification
3. Induction
● From observations → general laws
● “data speaks for itself”
👉 BUT:
● ❌ logically invalid (black swan problem)
● You can never be 100% sure
,4. Behaviourism (application)
● Only study observable behavior in response to external stimuli
● Ignore thoughts → not measurable, so unverifiable
👉 This is positivism applied to psychology
Popper (Alternative)
- Science = falsifiability > “Can we prove it wrong?”
- Not: “Can we prove it?”
- Deduction: Theory > test
Fallibility
- We can NEVER be certain
- Only: “not yet falsified”
Dogmatic vs. critical thinking
- Dogmatic thinking seeks confirmation, defends theories at all costs, and ignores
counter-evidence.
- Critical thinking welcomes criticism and treats refutation as progress. A good theory
is not one that is confirmed often, but one that can be tested and potentially proven
wrong.
Positivism Popper
Verification Falsification
Induction Testing & refuting
Certainty possible Always uncertain
Data → theory Theory → test
, Lecture 2
Science ought to be value-free > Values should not play any role ins scientific research
Kinds of values
1. Epistemic values:
- values that have to do with truth and knowledge: (influence is unproblematic)
- e.g., truth, knowledge, explanatory scope, predictive power, etc.
- Constitutive role: are necessary for the activity of research, shape it from the inside
- what method gives the most accurate results, which analysis technique is best
2. Non-epistemic values:
- values that don’t have to do with truth and knowledge:
- e.g., political ideals, social values, religious and other worldviews
- Contextual role: Involved in the context of the research but not necessary for the
conducting.
- what research to fund, what problems to investigate, where and how to communicate
about findings, with whom are results shared, etc.
- Problem: still play a constitutive role = problem
In het kort:
Epistemic = waarheid, kennis → OK
Non-epistemic = politiek, moraal → tricky
Moderate Thesis of Value-Freedom Science
Science is objective when only epistemic values are constitutive and moral and political
values always remain contextual (Optimistic)
Errors:
type I error (false positive)
- concluding there is some effect when there is none
type II error (false negative)
- concluding there is no effect when there is one
Values in science:
1. Weighing different kind of errors
2. Defining concepts
3. Methods and morality
4. Emancipatory and critical research
social science is value-laden in different ways
• This can be a problem, when values lead to bias: e.g., choosing a method because it
makes a desired outcome more likely, throwing out data points because they don’t fit a
preferred conclusion, interpreting conclusions in a one-sided manner, etc.
• But it need not be a problem, when social scientists are aware of
and transparent about value commitments, and don’t let them bias
methods, data gathering, analysis, and conclusions.
PoS: Investigating the social sciences
Social sciences: Investigating the social world
Demarcation problem: is about drawing a clear line between what counts as science and
what does not. > De grens is vaag tussen wat wel science en wat niet is
What makes something scientific? (demarcation problem)
And you have two main answers:
● Logical positivism → verification > Empirically verified
● Popper → falsification > Can we prove it wrong?
Logical positivism: A statement is only meaningful if it can be empirically verified.
- “Water boils at 100°C” → scientific
- “God exists” → not scientific (not testable)
Key elements Logical Positivism
1. Verifiability criterion
● Meaning = method of verification
● If you can’t test it → it’s meaningless
● To understand a statement = knowing how you would check if it’s true
2. Ideal language of science
● A statement only has meaning if you can test it in the real world.
● Use logic + math
● Make science precise and objective
● What gives a statement meaning in positivism?”
● Correct answer:
→ its method of verification
3. Induction
● From observations → general laws
● “data speaks for itself”
👉 BUT:
● ❌ logically invalid (black swan problem)
● You can never be 100% sure
,4. Behaviourism (application)
● Only study observable behavior in response to external stimuli
● Ignore thoughts → not measurable, so unverifiable
👉 This is positivism applied to psychology
Popper (Alternative)
- Science = falsifiability > “Can we prove it wrong?”
- Not: “Can we prove it?”
- Deduction: Theory > test
Fallibility
- We can NEVER be certain
- Only: “not yet falsified”
Dogmatic vs. critical thinking
- Dogmatic thinking seeks confirmation, defends theories at all costs, and ignores
counter-evidence.
- Critical thinking welcomes criticism and treats refutation as progress. A good theory
is not one that is confirmed often, but one that can be tested and potentially proven
wrong.
Positivism Popper
Verification Falsification
Induction Testing & refuting
Certainty possible Always uncertain
Data → theory Theory → test
, Lecture 2
Science ought to be value-free > Values should not play any role ins scientific research
Kinds of values
1. Epistemic values:
- values that have to do with truth and knowledge: (influence is unproblematic)
- e.g., truth, knowledge, explanatory scope, predictive power, etc.
- Constitutive role: are necessary for the activity of research, shape it from the inside
- what method gives the most accurate results, which analysis technique is best
2. Non-epistemic values:
- values that don’t have to do with truth and knowledge:
- e.g., political ideals, social values, religious and other worldviews
- Contextual role: Involved in the context of the research but not necessary for the
conducting.
- what research to fund, what problems to investigate, where and how to communicate
about findings, with whom are results shared, etc.
- Problem: still play a constitutive role = problem
In het kort:
Epistemic = waarheid, kennis → OK
Non-epistemic = politiek, moraal → tricky
Moderate Thesis of Value-Freedom Science
Science is objective when only epistemic values are constitutive and moral and political
values always remain contextual (Optimistic)
Errors:
type I error (false positive)
- concluding there is some effect when there is none
type II error (false negative)
- concluding there is no effect when there is one
Values in science:
1. Weighing different kind of errors
2. Defining concepts
3. Methods and morality
4. Emancipatory and critical research
social science is value-laden in different ways
• This can be a problem, when values lead to bias: e.g., choosing a method because it
makes a desired outcome more likely, throwing out data points because they don’t fit a
preferred conclusion, interpreting conclusions in a one-sided manner, etc.
• But it need not be a problem, when social scientists are aware of
and transparent about value commitments, and don’t let them bias
methods, data gathering, analysis, and conclusions.
