#internal-validity

Maximize internal validity through experimental control. Threats to internal validity (Campbell & Stanley): 1. History: external events during study. Control: randomization, brief study duration. 2. Maturation: natural changes over time. Control: control group, random assignment. 3. Testing: effects of pretesting. Control: Solomon four-group design, posttest-only design. 4. Instrumentation: changes in measurement. Control: standardized protocols, calibration. 5. Regression to mean: extreme scores regress toward average. Control: random assignment, cutoff-based assignment analysis. 6. Selection: systematic differences between groups. Control: randomization, matching. 7. Mortality: differential dropout. Control: intent-to-treat analysis, retention strategies. Design features: random assignment is gold standard. Manipulation checks ensure independent variable was successfully manipulated. Attention controls eliminate placebo effects.