Introduction & Research Methods

Before the five methods, get the logic clear — it underpins every Paper 2 answer.

A research method is a systematic way of collecting data to test a hypothesis or answer a question. Choosing a method is always a trade-off between control and realism.

Cause and effect (causation). To show that one thing causes another, you must change one variable (the independent variable, IV) and measure its effect on another (the dependent variable, DV) while controlling everything else. Only the experiment does this — so only experiments can establish cause and effect. Every other method describes relationships or behaviour but cannot prove cause.

Quantitative vs qualitative data.

• Quantitative data = numbers (e.g. how many helped, reaction time). Easy to analyse and compare; can lack depth.
• Qualitative data = descriptions/words (e.g. interview responses, observed comments). Rich and detailed; harder to analyse objectively.

Objectivity vs subjectivity. Objective data are free from personal bias (e.g. a stopwatch reading); subjective data depend on opinion/interpretation (e.g. rating how 'aggressive' a child looked). Scientific psychology aims to be objective.

Why this matters. When you evaluate any method in the exam, you are really commenting on how much control it has (→ cause and effect, reliability), how realistic it is (→ ecological validity), and what kind of data it gives (numbers vs depth). Hold those three lenses and you can evaluate any method.

The experiment manipulates an independent variable (IV) and measures its effect on a dependent variable (DV), controlling extraneous variables. It is the only method that shows cause and effect.

Laboratory experiment — conducted in a controlled environment.

• Strengths: high control of extraneous variables → confident cause and effect; standardised → replicable and reliable.
• Weaknesses: the artificial setting lowers ecological validity; can produce demand characteristics (participants guess the aim).
• Example: Milgram's obedience study (controlled lab setting).

Field experiment — the IV is manipulated in a real-life setting.

• Strengths: higher ecological validity (real setting, natural behaviour); often fewer demand characteristics if participants are unaware.
• Weaknesses: less control of extraneous variables (confounds); often raises ethical issues (consent, debriefing).
• Example: Piliavin's subway study (IV = victim type, manipulated on a real train).

The key contrast: lab = control (cause and effect, replicable) but artificial; field = realism (ecological validity) but harder to control.

Self-report = participants report their own thoughts, feelings or behaviour.

Questionnaires (written).

• Closed questions give fixed responses (e.g. yes/no, rating scales) → quantitative, easy to analyse and compare.
• Open questions allow free responses → qualitative, rich but harder to analyse.
• Strengths: can collect large amounts of data quickly and cheaply.
• Weaknesses: prone to social desirability bias (people answer to look good) and misunderstanding; low response/return rates.

Interviews (spoken).

• Structured — fixed questions in a set order → replicable, comparable, but inflexible.
• Unstructured — a conversation guided by responses → rich, detailed, but hard to replicate and analyse.
• Semi-structured — set questions plus follow-ups (a blend).
• Strengths: can probe for detail; clarify misunderstandings.
• Weaknesses: interviewer effects and social desirability; unstructured interviews are hard to compare.

Across all self-reports: the data depend on participants being honest and accurate — a key limitation.

Observation = watching and recording behaviour. You must be able to classify an observation along four dimensions.

1. Setting: naturalistic (in the everyday environment → high ecological validity, low control) vs controlled (in a structured setting → more control, less natural).

2. Researcher's role: participant (the observer joins the group → rich insight, but risks bias/losing objectivity) vs non-participant (observes from outside → more objective, but less insight).

3. Awareness: covert (participants don't know they're observed → natural behaviour, but ethical issues over consent) vs overt (they know → ethical, but risks demand characteristics).

4. Structure: structured observations use behavioural categories (clearly defined behaviours to tally) → objective, reliable, quantitative. Unstructured observations record everything → rich but harder to analyse.

Improving observations: clear behavioural categories and two or more observers allow a check of inter-observer reliability (do independent observers record the same thing?).

Strengths overall: capture real behaviour (especially naturalistic) without relying on self-report honesty. Weaknesses: observer bias; cannot explain why behaviour happens; ethical issues if covert.

Case study = a detailed, in-depth investigation of a single individual, small group or event, usually over time, using several methods (interviews, observations, tests, etc.) — so it combines techniques (triangulation).

Strengths:

• Rich, detailed (qualitative) data about a real case — high in depth and often ecological validity.
• Can study rare or unique cases that could not be created experimentally (e.g. brain damage).
• Using several techniques allows triangulation (cross-checking findings).

Weaknesses:

• Cannot generalise from one case to everyone (small, unique sample).
• Researcher bias / subjectivity in interpreting rich data.
• Often cannot be replicated (each case is unique), and may rely on retrospective (memory-based) data.

Example: Saavedra & Silverman's case study of a single boy's phobia.

Correlation measures the relationship (association) between two co-variables — it does not manipulate anything.

Types of correlation:

• Positive — as one variable increases, the other increases (e.g. more revision → higher marks).
• Negative — as one increases, the other decreases (e.g. more stress → less sleep).
• No correlation — no relationship.

Strength is shown by a correlation coefficient (from +1 through 0 to −1); the closer to ±1, the stronger the relationship. Correlations are usually displayed on a scatter graph.

The golden rule: correlation ≠ causation. A correlation cannot show cause and effect because:

• the direction of causation is unknown (does A cause B, or B cause A?), and
• a third (extraneous) variable could cause both.

Strengths: useful when manipulation is impractical or unethical; can reveal relationships worth investigating further (then by experiment). Weaknesses: no cause and effect; can be misinterpreted as causal.