Research methods: experiments and other methods, sampling, experimental design, variables and hypotheses, descriptive and inferential statistics, and the chosen inferential tests.

What this dot point is asking

Edexcel weaves research methods through every topic, so you must choose appropriate methods, design valid studies, handle data, and select and justify the correct inferential test from the Edexcel list. The skills here are tested across all three papers, especially in data-handling questions.

The answer

Methods, variables and hypotheses

Beyond the experiment (laboratory, field, natural and quasi), psychologists use observations (naturalistic or controlled, structured or unstructured), self-report (questionnaires with open and closed questions, structured and unstructured interviews), correlations (measuring the relationship between two co-variables) and case studies (in-depth study of one person or group). Each balances control against ecological validity.

Sampling and experimental design

Descriptive and inferential statistics

Descriptive statistics summarise data: measures of central tendency (mean, median, mode) and dispersion (range and standard deviation), plus graphs (bar charts, histograms, scattergrams) and tables. The standard deviation measures the average spread of scores around the mean.

Reliability, validity and ethics (GRAVE)

• Generalisability. A representative sample (favoured by random or stratified sampling) allows findings to generalise; biased samples (volunteer, opportunity) limit it.
• Reliability. Standardised procedures, test-retest and inter-rater checks make a study consistent and replicable.
• Application. Sound methods let psychology be applied with confidence, for example to design effective health campaigns or therapies.
• Validity. Internal validity (the IV, not a confound, caused the change) and ecological validity (findings apply outside the artificial setting) must both be considered.
• Ethics. The BPS code requires informed consent, no unnecessary deception, the right to withdraw, confidentiality and protection from harm, with debriefing where deception is used.

Examples in context

Example 1. Reading a chi-square ($\chi^2$) result. Suppose a study categorises participants by whether they were primed with a word (yes or no) and whether they then recalled a target (recalled or not). These are nominal frequency counts in an unrelated design, so chi-square is the correct test. With a $2 \times 2$ contingency table, the degrees of freedom are $df = (2 - 1)(2 - 1) = 1$, and the critical value at $p < 0.05$ is $3.84$. If the calculated $\chi^2 = 5.20$, then because $5.20 > 3.84$ the result is significant: there is a significant association between priming and recall, with less than a 5 per cent probability the result is due to chance. Note that chi-square is one of the few tests where the calculated value must be larger than the critical value.

Example 2. Spearman's rho for a correlation. A researcher measures hours of revision and exam mark for 20 students and finds Spearman's $rho = +0.78$ ($p < 0.05$). Spearman is the correct test because the data are ordinal and the aim is to measure a relationship rather than a difference. The coefficient $+0.78$ is a strong positive correlation: more revision is associated with higher marks. The significance ($p < 0.05$) means the relationship is unlikely to be chance, but it does not prove revision causes higher marks, because correlations cannot establish causation (a third variable such as motivation could drive both). This shows how the choice of test follows from the research aim and data type.

Try this

Q1. A study compares recall scores (ordinal data) between two separate groups. Name and justify the inferential test. [3 marks]

• Cue. Mann-Whitney U, because the data are ordinal and the design is unrelated (independent groups), testing for a difference.

Q2. Explain what is meant by a significant result at $p < 0.05$. [2 marks]

• Cue. There is a 5 per cent or less probability that the result is due to chance, so the null hypothesis can be rejected.

Q3. Explain the difference between a directional and a non-directional hypothesis, using an example. [4 marks]

• Cue. A directional (one-tailed) hypothesis states the direction of difference (group A recalls more than group B); a non-directional (two-tailed) hypothesis states only that there is a difference (group A and B differ in recall).