Observational vs Experimental Studies Math Example 4
Follow the full solution, then compare it with the other examples linked below.
Example 4
hardWhy is it impossible to conduct a fully randomized experiment to test the effect of poverty on health outcomes? What study design limitations result from this constraint?
Solution
- 1 Impossible to randomize poverty: we cannot ethically or practically assign people to poverty or wealth
- 2 Limitation 1: confounders cannot be fully controlled — poor health may cause poverty (reverse causation) as much as poverty causes poor health
- 3 Limitation 2: observational data (census, hospital records) suffers from selection bias and unmeasured confounders
- 4 Partial solutions: natural experiments (policy changes that affect some regions but not others); instrumental variables; regression discontinuity designs
Answer
Randomizing poverty is unethical. Results in observational studies that can show association but not definitively prove causation.
Many important social science questions (poverty, education, discrimination) cannot be studied experimentally for ethical or practical reasons. Researchers use quasi-experimental designs (natural experiments, IVs) to approximate experimental conditions from observational data.
About Observational vs Experimental Studies
An observational study records data without imposing treatments, while an experiment deliberately manipulates a variable. Only experiments with random assignment can establish causation; observational studies can only show association.
Learn more about Observational vs Experimental Studies →More Observational vs Experimental Studies Examples
Example 1 medium
Classify each study and state what conclusions can be drawn: (a) Researchers record which students e
Example 2 hardAn observational study finds smokers have 10x the lung cancer rate of non-smokers. Critics say this
Example 3 easyA researcher asks 500 adults about their exercise habits and mental health. Is this observational or