Summary Research Methodology Notes

Time order problems:
- When the variables are measured at the same time
- The variables change each other
- Reverse causation

Independent variable Dependent variable
X Y
Cause Effect
Exogenous Endogenous
Treatment Observation

Deterministic: if we do this, then this always happen
Probabilistic: if we do this, then probably this will happen

Cross sectional research design
- O
- Association
- Time order problems sometimes excluded
- Internal validity is weak because of the reverse causation and the third
variable
- Confounders
- Strong validity: a lot of units

Interrupted times series
- OXO
- Longitudinal research
- Internal validity is weak because third variable is not taking into account.
- External validity: is it valid for the whole population
- Measurement and sample validity: treatment related to the theoretical
constructs?
- Association
- Time order

Classical experiment
- ROXO
- RO O
- Third variable, association and time order
- Only for simple hypothesis

Research design: experiment, interrupted time series, cross sectional
Data collection: survey, observation, interview

Internal validity and statistical conclusion validity are about drawing the
correct conclusion in the research.
External validity and sample & measurement validity are about
generalization and inference to the whole population/other cases/theory

Statistical conclusion validity: is the established relationship/association
correct? Correct handling of the data & using appropriate tests.
Internal validity: time order and third variable?
Measurement and sample validity: measuring the variable in the right way
according to the theoretical constructs.