Foundations of Research Design
Core design concepts: variables, validity types, and the trade-offs between experimental and observational designs.
Slide 1 of 4
Variables and operationalization
- An independent variable is manipulated or grouped; a dependent variable is measured.
- Operationalization specifies the concrete procedure used to measure a construct.
- Confounds are third variables that offer a rival explanation for an observed effect.
Slide 2 of 4
The four validities
- Internal validity: can the effect be attributed to the IV rather than a confound?
- External validity: do the findings generalize across people, settings, and time?
- Construct validity: do the measures and manipulations represent the intended constructs?
- Statistical conclusion validity: are inferences about covariation sound?
Slide 3 of 4
Experimental vs. observational designs
- Random assignment supports causal claims by equating groups in expectation.
- Observational designs reveal association but are vulnerable to confounding.
- Quasi-experiments lack random assignment but use design features to reduce threats.
Slide 4 of 4
Common threats to internal validity
- History, maturation, and testing effects across repeated measurement.
- Regression to the mean when groups are selected on extreme scores.
- Attrition that is differential across conditions.