Philosophy
● Mother of all scientific disciplines
● Scientific disciplines gradually disengage from philosophy as they develop.
● Philosophy continues to study the foundations of sciences
(including ethics). It is a meta-science (“behind science”).
Philosophy of science: studies the foundations, methods and
implications of science.
An assumption is a statement accepted without proof
Assumptions of science
Fundamental assumptions
● The existence of a true (material) reality
● Reality is ordered
● The order of reality can be discovered
● The discovered order is never final
Based on data, we often make inferences about an invisible world: Constructs.
Examples: gravity, intelligence, Freud’s id
Reification of a construct: tendency to treat constructs as if they are part of reality.
Hypothesis = A testable explanation of a phenomenon – a mini-theory.
To test a hypothesis, a prediction must be derived from it,which is tailored to a specific
situation.
Note: In the book, the hypothesis and prediction are treated as identical, but we distinguish
them here.
- The hypothesis is a testable idea, but not directly observable
- The prediction is an observable consequence of a hypothesis, tied to a specific
situation
,Theory = System of logically coherent constructs and statements about a certain area of
reality.
● No contradictions among statements
● From a theory, (partial) hypotheses can be deduced
● Must be falsifiable and parsimonious
Parsimony: among competing theories, the simplest (most sparse) theory should be
preferred ( Ockham’s razor).
Test:
● Data inconsistent with prediction falsification: the hypothesis is not true
Test:
● Data consistent with prediction verification? Can we say the hypothesis is true?
→ There is a logical asymmetry between falsification and verification. Why?
Scientific argument
1. 2.
3.confirmation of the consequent 4.Denial of the antecedant
5. Denial of the consequent, Modus Tollens (MT) 6. Not H..en Confirmation of the consequent
,7. Better attempt: Verification
A good theory:
● offers many opportunities for falsification
● delivers unique and "risky" predictions (bold predictions)
Bold prediction
Hypothetical example
Predictions of climate theories A and B:
A. Next year, the summer will be warmer than the spring.
B. Next year, the summer will be colder than the spring.
What is more impressive?
● Prediction of theory A comes true
● Prediction of theory B comes true
Ideal situation:
● two strong theories lead to different predictions
● test these predictions (in ‘critical experiment’)
Summary
● A scientific hypothesis (or theory) should be falsifiable
● A hypothesis is especially valuable if its predictions:
- are surprising or bold (far from obvious)
- conflict with predictions from other hypotheses
● A hypothesis can be falsified when a prediction derived from it does not come true
● But a hypothesis cannot be verified (or proven to be right) when a prediction derived
from it is confirmed
Pseudoscience
Characteristics
● Postulation of unfalsifiable ideas
● Claims based on incidental (or biased) observation
● Resistance against scientific research
● Ignore/deny counter evidence
"Develop your hidden psychic powers!"
“Lose weight by hypnosis!"
"Learn French as you sleep!"
Demarcation criteria of science
1. Systematic empiricism
, - unbiased data collection
2. Focus on testable theories/hypotheses
- theory needs to be falsifiable
- data determine the fate of theory
3. Publicly accessible
- articles in peer reviewed journals
Summary
Explaining is the most ambitious goal of science
● by means of an hypothesis (H) or a (more complex) theory
● H must be falsifiable
– if H is wrong, one must be able to show this
● We test H by deriving a prediction (P) from it
● If P does not come true falsification of H (deductive argument)
● If P comes true no support for H if P is obvious (unsurprising) it supports H if P is
nonobvious, unique or “bold”
● We can never verify (or “proof”) H (inductive argument)
● Pseudoscience escapes rigorous scientific scrutiny
Measured versus Manipulated
Measured variable:
● Obtained by observing the participant. For example:
- IQ
- cortisol level in saliva (following a stress task)
Manipulated variable:
● Determined by the researcher. For example:
- Real drug versus placebo (pill without active substance)
- Condition with or without threat of
getting a shock
Operational definition
● Mother of all scientific disciplines
● Scientific disciplines gradually disengage from philosophy as they develop.
● Philosophy continues to study the foundations of sciences
(including ethics). It is a meta-science (“behind science”).
Philosophy of science: studies the foundations, methods and
implications of science.
An assumption is a statement accepted without proof
Assumptions of science
Fundamental assumptions
● The existence of a true (material) reality
● Reality is ordered
● The order of reality can be discovered
● The discovered order is never final
Based on data, we often make inferences about an invisible world: Constructs.
Examples: gravity, intelligence, Freud’s id
Reification of a construct: tendency to treat constructs as if they are part of reality.
Hypothesis = A testable explanation of a phenomenon – a mini-theory.
To test a hypothesis, a prediction must be derived from it,which is tailored to a specific
situation.
Note: In the book, the hypothesis and prediction are treated as identical, but we distinguish
them here.
- The hypothesis is a testable idea, but not directly observable
- The prediction is an observable consequence of a hypothesis, tied to a specific
situation
,Theory = System of logically coherent constructs and statements about a certain area of
reality.
● No contradictions among statements
● From a theory, (partial) hypotheses can be deduced
● Must be falsifiable and parsimonious
Parsimony: among competing theories, the simplest (most sparse) theory should be
preferred ( Ockham’s razor).
Test:
● Data inconsistent with prediction falsification: the hypothesis is not true
Test:
● Data consistent with prediction verification? Can we say the hypothesis is true?
→ There is a logical asymmetry between falsification and verification. Why?
Scientific argument
1. 2.
3.confirmation of the consequent 4.Denial of the antecedant
5. Denial of the consequent, Modus Tollens (MT) 6. Not H..en Confirmation of the consequent
,7. Better attempt: Verification
A good theory:
● offers many opportunities for falsification
● delivers unique and "risky" predictions (bold predictions)
Bold prediction
Hypothetical example
Predictions of climate theories A and B:
A. Next year, the summer will be warmer than the spring.
B. Next year, the summer will be colder than the spring.
What is more impressive?
● Prediction of theory A comes true
● Prediction of theory B comes true
Ideal situation:
● two strong theories lead to different predictions
● test these predictions (in ‘critical experiment’)
Summary
● A scientific hypothesis (or theory) should be falsifiable
● A hypothesis is especially valuable if its predictions:
- are surprising or bold (far from obvious)
- conflict with predictions from other hypotheses
● A hypothesis can be falsified when a prediction derived from it does not come true
● But a hypothesis cannot be verified (or proven to be right) when a prediction derived
from it is confirmed
Pseudoscience
Characteristics
● Postulation of unfalsifiable ideas
● Claims based on incidental (or biased) observation
● Resistance against scientific research
● Ignore/deny counter evidence
"Develop your hidden psychic powers!"
“Lose weight by hypnosis!"
"Learn French as you sleep!"
Demarcation criteria of science
1. Systematic empiricism
, - unbiased data collection
2. Focus on testable theories/hypotheses
- theory needs to be falsifiable
- data determine the fate of theory
3. Publicly accessible
- articles in peer reviewed journals
Summary
Explaining is the most ambitious goal of science
● by means of an hypothesis (H) or a (more complex) theory
● H must be falsifiable
– if H is wrong, one must be able to show this
● We test H by deriving a prediction (P) from it
● If P does not come true falsification of H (deductive argument)
● If P comes true no support for H if P is obvious (unsurprising) it supports H if P is
nonobvious, unique or “bold”
● We can never verify (or “proof”) H (inductive argument)
● Pseudoscience escapes rigorous scientific scrutiny
Measured versus Manipulated
Measured variable:
● Obtained by observing the participant. For example:
- IQ
- cortisol level in saliva (following a stress task)
Manipulated variable:
● Determined by the researcher. For example:
- Real drug versus placebo (pill without active substance)
- Condition with or without threat of
getting a shock
Operational definition
