Chapter 4. Research Methods in the Social Sciences

The research process: choose a topic, review existing literature, form a research question or hypothesis, design the study, collect data, analyse, interpret and report [31][32].

Quantitative methods collect numerical data. Examples are surveys with closed questions, experiments, and analysis of official statistics. They test hypotheses and measure patterns. Qualitative methods gather non-numerical data, such as interviews, focus groups, observation, ethnography and analysis of texts or images, to explore meanings and experiences in depth. Mixed methods combine both [31][32].

Sampling. The population is the whole group of interest; a sample is the part studied. Random sampling gives each member an equal chance of selection and supports generalization; convenience sampling uses whoever is easy to reach and may be biased. Larger, representative samples reduce error.

Good measurement

Validity: does the instrument measure what it claims to?

Reliability: does it give consistent results?

Questions should be clear, neutral and free of leading wording.

Correlation and causation. Two variables may move together without one causing the other; a third factor (confounder) may be responsible. Experiments with random assignment provide the strongest evidence for causation [31]. Example: ice-cream sales and drowning both rise in summer, but ice cream does not cause drowning.

Research ethics. Respect for persons, beneficence (do good, avoid harm) and justice are guiding principles. In practice: informed consent, voluntary participation, confidentiality, protection of vulnerable people and honest reporting [33]. Studies such as Milgram’s obedience experiments led to stricter ethical review [16][33].

Reading statistics. Check sample size, how groups were chosen, margin of error, averages versus spread, and the difference between statistically significant and practically important.

Review: Why does a random sample help researchers generalize? (Answer: it gives every member of the population an equal chance of selection, reducing bias.)