Research methodology

· Research methodology threads through the whole DP Psychology course and is organised into research methods, sampling techniques, procedures, data collection, and data analysis/interpretation.
· It explains how psychologists make choices about how to investigate behaviour, how to measure variables, and how to draw conclusions from data.
· Key syllabus concepts most strongly linked to this topic: bias, causality, measurement, responsibility, validity, and reliability.
· Research methodology is studied within all contexts: health and well-being, human development, human relationships, and learning and cognition.
· High-scoring answers should connect methods to the research question, ethical considerations, validity/reliability, bias, generalizability, and whether causality can be inferred.

This page gives a clean overview of major psychological research approaches and their strengths and weaknesses. It is useful for revising how different research methods answer different types of questions. Use it to support the distinction between descriptive/correlational methods and experimental methods. Source

Research designs and research methods

· Experiments: include true experiments and quasi-experiments; useful when investigating possible cause-and-effect relationships.
· True experiments: researchers manipulate an independent variable (IV) and measure a dependent variable (DV), usually with stronger control over extraneous variables.
· Quasi-experiments: use naturally occurring groups or conditions; useful when true manipulation is impossible or unethical, but causality is usually weaker.
· Observations: can be naturalistic or controlled, overt or covert, participant or non-participant; useful for studying behaviour as it occurs.
· Surveys/questionnaires: collect self-report data from larger samples; useful for attitudes, beliefs, experiences, and prevalence-type questions.
· Interviews: may be structured, semi-structured, or focus group; useful for detailed qualitative data and participant perspectives.
· Correlational studies: examine the relationship between two variables; they can show association, but correlation does not prove causation.
· Case studies: intensive study of one person, group, event, or institution; useful for rare or complex behaviour but often limited in generalizability.

Choosing an appropriate method

· Match the method to the aim and research question: causality → experiment; relationship → correlation; lived experience → interview; behaviour in context → observation; large-scale patterns → survey.
· Evaluate each method by its advantages and disadvantages, not by memorising a fixed “best method”.
· Consider whether the method allows enough control, ecological validity, standardisation, and replicability.
· Consider whether the method produces quantitative data, qualitative data, or both.
· Link method choice to ethical considerations, especially when studying sensitive topics, children, deception, public behaviour, or vulnerable groups.

Sampling techniques

· Self-selected sampling: participants volunteer; practical and quick, but may produce volunteer bias.
· Opportunity sampling: uses people who are available; efficient but often weak for representativeness.
· Stratified sampling: selects participants from key subgroups in proportion to the target population; improves representativeness but requires population information.
· Random sampling: each member of the target population has an equal chance of selection; reduces sampling bias but can be difficult to organise.
· Snowball sampling: existing participants recruit others; useful for hard-to-reach groups but can create a narrow, network-based sample.
· Always evaluate sampling using representativeness, sampling bias, generalizability, and practical/ethical feasibility.

This diagram classifies sampling methods and helps students distinguish probability from non-probability approaches. It is useful for linking sampling choices to representativeness, bias, and generalizability. Use it when revising random, stratified, opportunity/self-selected, and snowball sampling. Source

Research considerations: validity, reliability, credibility

· Reliability: whether a measure or procedure gives consistent results.
· Validity: whether a method or measure actually measures what it claims to measure.
· Internal validity: whether changes in the DV are likely caused by the IV rather than extraneous variables.
· External validity: whether findings can apply beyond the study setting, sample, and procedure.
· Content validity: whether a measure covers the full range of the construct being studied.
· Face validity: whether a measure appears, on the surface, to measure the intended construct.
· Construct validity: whether the operationalised measure genuinely represents the psychological construct.
· Generalizability: whether results can be applied to another population or context.
· Transferability: whether qualitative findings can be meaningfully applied to another context after considering similarities and differences.
· Credibility: whether qualitative findings are believable and well-supported, often strengthened by careful procedures and checking interpretations.
· Reflexivity: researchers actively reflect on their own assumptions, background, and possible unconscious bias.

The bullseye diagram shows that reliability means consistency, while validity means accuracy. A measure can be reliable without being valid, so exam answers should not treat the two terms as interchangeable. This image is especially useful for explaining measurement quality in research methodology. Source

Bias and controls

· Researcher bias: the researcher’s expectations or assumptions influence design, data collection, interpretation, or reporting.
· Participant bias: participants alter behaviour because they know they are being studied or want to respond in a certain way.
· Sampling bias: the sample does not represent the target population.
· Confirmation bias: researchers favour evidence that supports their expectations.
· Publication bias: significant or positive findings are more likely to be published than non-significant findings.
· Controls may include standardised procedures, clear operationalisation, control of extraneous variables, anonymity, inter-rater reliability, triangulation, and reflexivity.
· Strong evaluation explains both the source of bias and the effect on conclusions.

Causality and variables

· Causality means one variable produces a change in another, but human behaviour is often caused by interactions between multiple variables.
· Experiments are strongest for testing causality because they allow manipulation of an IV and measurement of a DV.
· Extraneous variables can weaken causal conclusions if they are not controlled.
· Correlational studies cannot prove causation because of bidirectional ambiguity and possible third variables.
· Evaluation should distinguish influence, interaction, association, and cause-and-effect.

Data collection and types of data

· Quantitative data: numerical data; useful for graphs, descriptive statistics, inferential statistics, and comparisons.
· Qualitative data: non-numerical data such as interview transcripts or observation notes; useful for meaning, experience, and context.
· Self-reported data: data participants provide about themselves; may increase access to inner thoughts but can be affected by memory, social desirability, and demand characteristics.
· Empirical data: data based on observation or measurement.
· Operationalisation: defining a variable in measurable terms; essential for valid and reliable measurement.
· Triangulation: using more than one source, method, or perspective to strengthen credibility.

Data analysis and interpretation

· Data can be represented using bar graphs, box and whisker plots, frequency tables, histograms, line graphs, and scatterplots.
· Students should interpret distributions, including normal distribution, skewness, and outliers.
· Descriptive statistics summarise data: mean, median, mode, range, standard deviation, and semi-interquartile range.
· Inferential statistics help decide whether results are likely due to chance.
· Tests of difference may include chi-square, related t-test, unrelated t-test, Mann-Whitney test, and Wilcoxon test.
· Tests of relationship include correlation coefficients.
· Statistical significance indicates whether a result is unlikely to be due to chance, while effect size indicates the size or practical importance of an effect.
· Type I error: falsely concluding there is an effect when there is not.
· Type II error: failing to detect an effect when one does exist.
· HL note: data analysis and interpretation is studied by SL and HL, but not assessed at SL.

These figures show how different kinds of data are represented visually. They are useful for practising interpretation of frequency, distribution shape, skew, spread, and outliers. Use them when revising which graph type fits which data type. Source

Thematic analysis

· Thematic analysis is used to analyse textual qualitative data by identifying patterns and grouping them into themes relevant to the research aim.
· Key stages: familiarisation with data, initial coding, searching for themes, reviewing themes, defining/naming themes, and reporting findings.
· Themes should be linked clearly to the research question and supported by evidence from the data.
· Researcher reflexivity is important because interpretation can be influenced by assumptions and perspective.
· Thematic analysis is useful for interviews, focus groups, and open-ended qualitative responses.

This diagram shows the six main phases of thematic analysis from familiarisation to producing the report. It helps students remember that qualitative analysis is systematic, not just impression-based. Use it to revise how interview or focus-group data can be coded and organised into themes. Source

Ethics and researcher responsibility

· Ethical standards are part of researcher responsibility and affect both participants and research outcomes.
· Core ethical considerations: informed consent, right to withdraw, protection from harm, anonymity, confidentiality, debriefing, cost-benefit analysis, and careful use of deception.
· Ethical standards can affect results: informed consent may create participant variables; right to withdraw may affect sampling; anonymity may increase validity of self-report.
· Research involving children, animals, public spaces, or socially sensitive topics requires especially careful ethical evaluation.
· Psychologists should avoid stigmatising marginalised groups and communicate uncertainty in findings responsibly.

Class practicals

· All students must complete one class practical for each context.
· Health and well-being: recommended practical = interview; minimum sample = 1 individual interview participant or 3–8 focus group participants.
· Human development: recommended practical = observation; minimum sample = 1 participant.
· Human relationships: recommended practical = survey/questionnaire; minimum sample = 10 participants.
· Learning and cognition: recommended practical = experiment; minimum sample = 5 participants for repeated measures or 10 participants for independent measures.
· Students should record the aim, procedure, sampling technique, sample findings, and ethical considerations.
· All class practicals must follow ethical guidelines; studies creating anxiety, stress, pain, or discomfort are not permitted.
· Conformity or obedience studies are not permitted under any circumstances.
· Animals must not be used for class practicals.

Exam evaluation prompts

· For any study, ask: What method was used and why was it appropriate?
· Evaluate whether the study supports causality, association, or only description.
· Comment on sampling technique, sample size, and generalizability.
· Comment on validity, reliability, bias, and ethical considerations.
· Use precise language: “This limits internal validity because…”, “This weakens generalizability because…”, “This increases credibility because…”