Research Methods and Data Analysis
What this domain covers
This domain tests the research process end to end: turning a concept into a measurable variable, choosing a design and sample, judging validity and reliability, reading data, and spotting ethical violations. You will also be asked to interpret correlation and to recognize when a third variable explains an apparent relationship.
Core rules and concepts
- Operationalization assigns measurable indicators to an abstract concept (for example, social class as income plus education plus occupation).
- The independent variable does the explaining; the dependent variable is explained. A hypothesis is a testable statement of relationship.
- Reliability means consistency across repetitions; validity means measuring what you claim. A scale can be reliable and still invalid.
- Sampling: probability methods (simple random, systematic, stratified, cluster) permit generalization; nonprobability methods (convenience, snowball, quota) do not. A large sample never repairs a biased sampling frame.
- Methods: survey (breadth, self-report bias), experiment (causal leverage, artificiality), field research (depth, reactivity, limited generalizability), secondary analysis (inexpensive, constrained by existing data), content analysis, historical-comparative work.
- Correlation is not causation. A spurious relationship fades once a control variable is introduced.
- Ethics: informed consent, voluntary participation, confidentiality, no harm. Tuskegee, Milgram, Humphreys, and Zimbardo drove modern IRB review.
Worked example
Ice cream sales and drowning deaths rise together across the year, and a student concludes ice cream causes drowning. The link is spurious: hot weather drives both. Holding temperature constant weakens the association.
Traps that cost marks
- Declaring a measure invalid merely because it is unreliable, or the reverse.
- Accepting a huge convenience sample as representative.
- Reading an observational study as proof of cause.
- Forgetting reactivity: people change when they know they are observed.