Overview
The lecture compares observational studies and controlled experiments as data collection methods, focusing on how they determine relationships like association or causation, and highlights ethical considerations in research design.
Data Collection Methods
- Researchers use two main data collection methods: observational studies and controlled experiments.
- In observational studies, subjects choose their actions; researchers only observe and record.
- In controlled experiments, subjects are assigned specific actions or treatments by the researcher.
Case Study Comparison
- Example A: Participants were asked about their tea-drinking habits and bedtime (observational study).
- The keyword "asked" indicates subjects chose their actions themselves.
- Observational studies can only establish an association between the variables.
- Example B: Participants were randomly assigned to tea or no-tea groups (controlled experiment).
- The keyword "told" indicates subjects were assigned their actions.
- Controlled experiments can establish causation between variables.
Association vs. Causation
- Observational studies show associations (relationships without proof that one variable causes the other).
- Controlled experiments can establish causation (proving one variable directly affects another).
Ethics in Research
- Some controlled experiments (e.g., asking pregnant women to smoke) are unethical due to potential harm.
- Observational studies are used when controlled experiments are not ethical or feasible.
Key Terms & Definitions
- Observational Study — Data collection where subjects choose their own actions; only associations can be established.
- Controlled Experiment — Data collection where subjects are assigned actions by the researcher; causation can be established.
- Association — A relationship between two variables without clear evidence that one causes the other.
- Causation — A relationship where one variable is proven to directly cause a change in another variable.
- Ethics — Considerations of right and wrong that influence which research methods are appropriate.
Action Items / Next Steps
- Practice identifying data collection methods and relationships (association vs. causation) in different study scenarios.
- Review ethical considerations when designing research studies.