Economic Sociology
Chapter 1 – Status of Sociology NOTES
Science vs. Pseudoscience
|| Characteristics of Science
Science ensures logical coherence and the absence of
contradictions.
Internal
Concepts and arguments must be clearly defined and open to
Consistency
falsification (= the ability to be proven false)
Use of structured language (e.g.: math) for consistency
Theories are supported by empirical data
Theories are tested through experiments
Dynamic interactions exist between theory and evidence
through continuous (scientific) cycle
Evidence The Scientific Cycle
Induction Deriving general principles from observations
Formulating explanations based on observed
Theory
patterns
Deriving specific predictions from general
Deduction principles
→ = Deducing hypothesis
Evidence Testing predictions with empirical data
Science does not eliminate but reduce errors over time
Reduction
Scientific progress is about the direction of ROE rather than
of Error
avoiding the errors entirely
(ROE)
Science does not validate data based on their age → ≠ pseudo
2 Levels of Social Organization of Science
Social Org. Scientific progress depends on an individual
Individual
of Science willingness to accept refutation
Science thrives on a social system of
Social competition and free exchange of ideas, similar
to a market economy
1
, Case – Trofim Lysenko
In the Soviet Union, Trofim Lysenko falsely claimed that plants
could inherit acquired traits. His theories, despite lacking
scientific basis, were politically endorsed, leading to the
suppression of genuine science and persecution of scientists
Can Sociology Be Scientific?
|| Natural Science vs. Sociology: 2 Fundamental Differences
If sociologist do their job
1. Limited use of experiments - Unlike physics or biology, sociology well, then yes it can be
cannot always rely on controlled experiments scientific
2. Focus on human interactions - Sociology studies social behaviour,
which is dynamic and context-dependent
|| The Experiment Design: OXO Model
3 Fundamentals Structures /!\ experimental group
always gets attributed
Randomized Participants are split into an experimental group something (e.g.
Distribution (receives treatment) and a control group (no treatment) medicine)
Treatment The experimental group undergoes a specific
The other group gets
Application intervention
‘placebo’ (control group)
Pre and Post-Test Observations are recorded before and after treatment to
Measurements evaluate the effect You then observe.
OXO model is ideal for testing causal relationships since it isolates the effects
of an intervention
Example:
Homelessness and drug use – they are addicted to alcohol, etc., if you compare
these samples of homeless people when it comes to addiction with the rest of Example:
the population it multiplies a lot Civil war – your
There is a relationship. leadership style starts
Question: changing after one – two
Do you become addicted because you’re homeless? generations in the
Or if you become addicted you lose your income? organisations. It shifts.
3 Strengths of Experiment Solomon design = four
Reliable and valid conclusions group design; combines
pretest-post-test and
Isolates the effect of an intervention, reducing external influences post-test-only designs to
ensure treatment effects
Several experimental variations exist (e.g.: Solomon designs and placebo- are not caused by the
controlled studies) pretest itself
2
,|| Challenges and Alternatives for Sociology
Experiments are difficult to simulate human behaviour in a laboratory setting.
Therefore, sociologists rely on large-scale surveys and quasi-experiments
Use standardized interviews to gather large
Charact. datasets. Multiple variables are collected to
identify patterns Second best options for
experiment designs
Surveys lack experimental control, making it
difficult to establish causation or leading to
spuriousness
Types of Spuriousness
Confounder A third variable that explains Spuriousness =
(common the observed relationship Relationships that occur
causation) between 2 variables when two variables
Purely coincidental appear related due to a
Random relationship between 2 confounder or a
Association variables that lacks logic or coincidence but are not
causal basis causally connected
Switched interpretation of the
Reverse “cause” and the “effect” within Example: nr of ice-
Causality a relationship between 2 cream sold vs. nr of
Large – Scale
variables drownings at seaside →
Surveys
Both variables in a relationship caused by good weather
Errors Feedback influence each other in a Cause: common
Loops circular manner causation = both things
→ = Bidirectional Causality caused by the third
variable
A relationship between 2
Structural variables appearing to be valid
Spurious. because the data is structured
or grouped
Spuriousness Reduction Strategies
Controlling confounding
Multivariate factors by including them into
Designs the research analysis as
independent variables
Observing the same
Longitudinal
individuals or groups over
Designs
extended periods of time
.
3
, A research design that resembles an experiment
but lacks random assignment of participants to
treatment and control groups and which occurs
in real-world settings where strict experimental
Definition
conditions are not possible
→ Does not fully meet the idea conditions of an
experiment
Comparison Groups – Instead of strict control
groups, quasi-experiments use naturally
occurring comparison groups
Quasi –
Experiments Natural Setting – Often conducted in real-world
environments, making findings more applicable
to actual societal conditions
Charact. No Randomization – Participants are not
randomly assigned to experimental and control
groups. Instead, groups may be pre-existing Example: 1854 London
Cholera Outbreak –
Limited Control Over Confounding Variables – “maybe it has something
Because groups are not randomly assigned, to do with the
researchers must use statistical techniques contaminated water”, he
(e.g., multivariate analysis) to reduce looked at where they got
spuriousness infected and where not,
it was traced back to the
water supply
Behaviour, action and social action
|| What is Sociology?
Definition The science of social phenomena made up of social actions
Sociology vs. Sociology is different from natural sciences as it studies
Nat. Science conscious, sentient beings rather than physical laws
Research
Simply asking people about their beliefs and actions
Strat.
|| Human Actions in Sociology
Human actions are not governed by fixed natural laws
Explanations on human actions are based on reasons rather
Key Ideas than just causality (e.g.: beliefs, values, motives))
Sociology does less (in experimentation) but also more (in
analyzing subjective human experiences) than nat. sciences
4
Chapter 1 – Status of Sociology NOTES
Science vs. Pseudoscience
|| Characteristics of Science
Science ensures logical coherence and the absence of
contradictions.
Internal
Concepts and arguments must be clearly defined and open to
Consistency
falsification (= the ability to be proven false)
Use of structured language (e.g.: math) for consistency
Theories are supported by empirical data
Theories are tested through experiments
Dynamic interactions exist between theory and evidence
through continuous (scientific) cycle
Evidence The Scientific Cycle
Induction Deriving general principles from observations
Formulating explanations based on observed
Theory
patterns
Deriving specific predictions from general
Deduction principles
→ = Deducing hypothesis
Evidence Testing predictions with empirical data
Science does not eliminate but reduce errors over time
Reduction
Scientific progress is about the direction of ROE rather than
of Error
avoiding the errors entirely
(ROE)
Science does not validate data based on their age → ≠ pseudo
2 Levels of Social Organization of Science
Social Org. Scientific progress depends on an individual
Individual
of Science willingness to accept refutation
Science thrives on a social system of
Social competition and free exchange of ideas, similar
to a market economy
1
, Case – Trofim Lysenko
In the Soviet Union, Trofim Lysenko falsely claimed that plants
could inherit acquired traits. His theories, despite lacking
scientific basis, were politically endorsed, leading to the
suppression of genuine science and persecution of scientists
Can Sociology Be Scientific?
|| Natural Science vs. Sociology: 2 Fundamental Differences
If sociologist do their job
1. Limited use of experiments - Unlike physics or biology, sociology well, then yes it can be
cannot always rely on controlled experiments scientific
2. Focus on human interactions - Sociology studies social behaviour,
which is dynamic and context-dependent
|| The Experiment Design: OXO Model
3 Fundamentals Structures /!\ experimental group
always gets attributed
Randomized Participants are split into an experimental group something (e.g.
Distribution (receives treatment) and a control group (no treatment) medicine)
Treatment The experimental group undergoes a specific
The other group gets
Application intervention
‘placebo’ (control group)
Pre and Post-Test Observations are recorded before and after treatment to
Measurements evaluate the effect You then observe.
OXO model is ideal for testing causal relationships since it isolates the effects
of an intervention
Example:
Homelessness and drug use – they are addicted to alcohol, etc., if you compare
these samples of homeless people when it comes to addiction with the rest of Example:
the population it multiplies a lot Civil war – your
There is a relationship. leadership style starts
Question: changing after one – two
Do you become addicted because you’re homeless? generations in the
Or if you become addicted you lose your income? organisations. It shifts.
3 Strengths of Experiment Solomon design = four
Reliable and valid conclusions group design; combines
pretest-post-test and
Isolates the effect of an intervention, reducing external influences post-test-only designs to
ensure treatment effects
Several experimental variations exist (e.g.: Solomon designs and placebo- are not caused by the
controlled studies) pretest itself
2
,|| Challenges and Alternatives for Sociology
Experiments are difficult to simulate human behaviour in a laboratory setting.
Therefore, sociologists rely on large-scale surveys and quasi-experiments
Use standardized interviews to gather large
Charact. datasets. Multiple variables are collected to
identify patterns Second best options for
experiment designs
Surveys lack experimental control, making it
difficult to establish causation or leading to
spuriousness
Types of Spuriousness
Confounder A third variable that explains Spuriousness =
(common the observed relationship Relationships that occur
causation) between 2 variables when two variables
Purely coincidental appear related due to a
Random relationship between 2 confounder or a
Association variables that lacks logic or coincidence but are not
causal basis causally connected
Switched interpretation of the
Reverse “cause” and the “effect” within Example: nr of ice-
Causality a relationship between 2 cream sold vs. nr of
Large – Scale
variables drownings at seaside →
Surveys
Both variables in a relationship caused by good weather
Errors Feedback influence each other in a Cause: common
Loops circular manner causation = both things
→ = Bidirectional Causality caused by the third
variable
A relationship between 2
Structural variables appearing to be valid
Spurious. because the data is structured
or grouped
Spuriousness Reduction Strategies
Controlling confounding
Multivariate factors by including them into
Designs the research analysis as
independent variables
Observing the same
Longitudinal
individuals or groups over
Designs
extended periods of time
.
3
, A research design that resembles an experiment
but lacks random assignment of participants to
treatment and control groups and which occurs
in real-world settings where strict experimental
Definition
conditions are not possible
→ Does not fully meet the idea conditions of an
experiment
Comparison Groups – Instead of strict control
groups, quasi-experiments use naturally
occurring comparison groups
Quasi –
Experiments Natural Setting – Often conducted in real-world
environments, making findings more applicable
to actual societal conditions
Charact. No Randomization – Participants are not
randomly assigned to experimental and control
groups. Instead, groups may be pre-existing Example: 1854 London
Cholera Outbreak –
Limited Control Over Confounding Variables – “maybe it has something
Because groups are not randomly assigned, to do with the
researchers must use statistical techniques contaminated water”, he
(e.g., multivariate analysis) to reduce looked at where they got
spuriousness infected and where not,
it was traced back to the
water supply
Behaviour, action and social action
|| What is Sociology?
Definition The science of social phenomena made up of social actions
Sociology vs. Sociology is different from natural sciences as it studies
Nat. Science conscious, sentient beings rather than physical laws
Research
Simply asking people about their beliefs and actions
Strat.
|| Human Actions in Sociology
Human actions are not governed by fixed natural laws
Explanations on human actions are based on reasons rather
Key Ideas than just causality (e.g.: beliefs, values, motives))
Sociology does less (in experimentation) but also more (in
analyzing subjective human experiences) than nat. sciences
4
