How do psychologists design valid, reliable and ethical studies?
Research methods (Section B): hypotheses, variables, experimental and non-experimental methods, experimental designs, sampling, reliability, validity, and ethics.
A focused answer to WJEC A-Level Psychology Unit 2 Section B on research methods: hypotheses, variables, experimental and non-experimental methods, experimental designs, sampling techniques, reliability, validity and ethics.
Reviewed by: AI editorial process; not yet individually human-reviewed
Have a quick question? Jump to the Q&A page
Jump to a section
What this dot point is asking
Unit 2 Section B examines research methods. You must understand hypotheses, variables, the main experimental and non-experimental methods, experimental designs, sampling techniques, reliability, validity, and ethics. These skills are tested again, in greater depth, in Unit 4, so a firm foundation here pays off twice.
Hypotheses and variables
A hypothesis is a precise, testable prediction. A directional (one-tailed) hypothesis states the direction of the difference or relationship ("group A will recall more than group B"); a non-directional (two-tailed) hypothesis predicts a difference or relationship without direction. The null hypothesis predicts no difference or relationship (any result is due to chance). Use a directional hypothesis when previous research points one way.
The independent variable (IV) is what the researcher manipulates; the dependent variable (DV) is what is measured. Both must be operationalised (defined in measurable terms, for example "recall measured as words correct out of 20"). Extraneous variables are nuisance variables that could affect the DV, and a confounding variable is one that varies systematically with the IV and so spoils the conclusion; both must be controlled.
Methods
Experimental designs
- Independent groups. Different participants in each condition. Avoids order effects but introduces participant variables; needs more participants.
- Repeated measures. The same participants in every condition. Controls participant variables but risks order effects (practice or fatigue), reduced by counterbalancing.
- Matched pairs. Different participants matched on key variables (such as age or IQ). Reduces participant variables without order effects, but matching is time-consuming and never perfect.
Sampling
The target population is the group the researcher wants to study; a sample is drawn from it and should be representative so findings generalise.
- Random sampling. Every member has an equal chance (for example names drawn from a list). Low bias, but needs a full list.
- Opportunity sampling. Whoever is available and willing. Quick and cheap, but often unrepresentative and prone to researcher bias.
- Volunteer (self-selected) sampling. People respond to an advert. Easy to gather, but volunteers may be atypical.
- Systematic sampling. Every nth person from a list. Objective, but not truly random.
- Stratified sampling. The sample reflects the proportions of subgroups in the population. Representative, but time-consuming.
Reliability and validity
Reliability is consistency. Internal reliability is consistency within a measure (checked by the split-half method); external reliability is consistency over time (checked by test-retest); inter-rater reliability is agreement between observers. Improve reliability with standardised procedures and clearly operationalised measures.
Validity is whether a study measures what it intends to. Internal validity is whether the IV (not a confound) caused the change in the DV. External validity includes ecological validity (generalising to real settings), population validity (generalising to other people) and temporal validity (generalising over time). Threats include demand characteristics and investigator effects.
Ethics
Research must follow ethical guidelines: informed consent, avoidance of deception (or full debriefing if used), protection from harm, confidentiality, and the right to withdraw. The researcher weighs the costs and benefits of a study, and special care is taken with vulnerable participants. Ethics is examined throughout the course, including in the contemporary debates and controversies.
Examples in context
Example 1. Counterbalancing a repeated-measures design. If everyone did the quiet condition first, practice would inflate the noisy-condition scores. Splitting the sample so half do quiet first and half do noisy first (counterbalancing) controls these order effects.
Example 2. Choosing a sampling method. A study needing strong generalisation to a school might use stratified sampling to match the proportions of year groups, whereas a quick pilot might use opportunity sampling and accept its bias.
Try this
Q1. What is an operationalised variable? [1 mark]
- Cue. A variable defined in precise, measurable terms (for example recall measured as words correct out of 20).
Q2. Identify the experimental design in which the same participants take part in every condition, and state one way to control its main weakness. [3 marks]
- Cue. Repeated measures; control order effects with counterbalancing.
Q3. Explain the difference between reliability and validity. [4 marks]
- Cue. Reliability is consistency (the same result on repetition); validity is accuracy (measuring what you intend, such as ecological validity for real settings). A measure can be reliable but not valid.
Exam-style practice questions
Practice questions written in the style of WJEC exam questions on this dot point, with worked answer explainers. The year tag is the paper they imitate, not the source.
WJEC specimen4 marksWrite a suitable directional (one-tailed) hypothesis for a study comparing recall in a quiet room with recall in a noisy room, and identify the independent and dependent variables.Show worked answer →
WJEC rewards an operationalised hypothesis and correctly identified variables.
A directional hypothesis: "Participants who learn a 20-word list in a quiet room will recall more words than participants who learn it in a noisy room."
The independent variable is the noise level of the room (quiet versus noisy), which is manipulated.
The dependent variable is the number of words correctly recalled out of 20, which is measured.
Markers reward an operationalised, directional statement and correct IV and DV. A non-directional version would simply predict a difference without stating the direction.
WJEC specimen6 marksExplain one strength and one weakness of using opportunity sampling, and describe how random sampling differs.Show worked answer →
WJEC rewards clear definitions and a developed comparison.
Opportunity sampling takes whoever is available and willing at the time. Strength: it is quick, cheap and convenient. Weakness: it is often unrepresentative and prone to researcher bias, because the people nearby may share characteristics, so findings may not generalise.
Random sampling gives every member of the target population an equal chance of selection, for example by drawing names from the population list. This reduces bias and improves representativeness, but it requires a full list of the population and people may still refuse to take part.
A strong answer defines each technique and explains, not just states, the strength and weakness.
Related dot points
- Contemporary debates (Section A): applying the five approaches to a current debate such as the ethics of neuroscience, the importance of mothering, conditioning children, the reliability of eyewitness testimony, and the value of positive psychology.
A focused answer to WJEC A-Level Psychology Unit 2 Section A on contemporary debates: how the five approaches inform current debates such as the ethics of neuroscience, the importance of mothering, conditioning children, eyewitness reliability and the value of positive psychology, and how to argue both sides.
- Core research for Unit 2: Milgram (1963) on obedience to authority and Kohlberg (1968) on the stages of moral development, including procedure, findings, conclusions and evaluation.
A focused answer to the WJEC A-Level Psychology Unit 2 core research: Milgram (1963) on obedience to authority and Kohlberg (1968) on moral development, covering procedure, findings, conclusions and evaluation of each study.
- Personal investigations (Section A): the practical investigations carried out across the course (such as a Stroop or memory experiment, an observation and a correlation) and the design, analysis and evaluation skills they assess.
An overview of the WJEC A-Level Psychology Unit 4 personal investigations: the practical investigations carried out across the course and the design, analysis, reporting and evaluation skills they assess, examined through questions in the Unit 4 paper.
- Application of research methods (Section B): brain-scanning techniques, longitudinal and cross-sectional studies, extended reliability and validity, distributions and descriptive statistics, and choosing and interpreting inferential tests.
A focused answer to WJEC A-Level Psychology Unit 4 Section B on the application of research methods: brain-scanning techniques, longitudinal and cross-sectional studies, extended reliability and validity, distributions and descriptive statistics, and choosing and interpreting inferential tests.
- The cognitive approach: assumptions, application to the formation of relationships, the therapy of cognitive behaviour therapy, the classic study of Loftus and Palmer (1974), and evaluation.
A focused answer to WJEC A-Level Psychology Unit 1 on the cognitive approach: its assumptions, its application to the formation of relationships, cognitive behaviour therapy, the classic study of Loftus and Palmer (1974), and how to evaluate the approach.
Sources & how we know this
- WJEC GCE Psychology specification (from 2015) — WJEC (2015)