Component 2: Methodologies (knowledge, understanding and evaluation of) — Eduqas A-Level Psychology

Component 2: Methodologies is a core component of Eduqas A-Level Psychology that equips students with the scientific skills needed to design, conduct, and evaluate psychological research. It covers the entire research process, from formulating hypotheses and operationalising variables to selecting sampling methods, experimental designs, and data collection techniques. Students also learn how to analyse data using descriptive and inferential statistics, and how to assess reliability, validity, and ethical considerations.

This component is essential because it underpins all other areas of psychology; every topic studied in Components 1 and 3 relies on methodological understanding. It develops critical thinking and the ability to evaluate research rigorously, which is highly valued in higher education and professional contexts. Mastery of methodologies also directly supports the skills assessed in the Independent Investigation (Component 4), where students apply these concepts to their own research project.

• →Variables: independent variable (IV), dependent variable (DV), extraneous variables, confounding variables, and operationalisation of variables.
• →Experimental designs: independent measures, repeated measures, and matched pairs, including their strengths, weaknesses, and control of order effects.
• →Sampling techniques: random, opportunity, volunteer, and stratified sampling, with evaluation of representativeness and generalisability.
• →Data analysis: descriptive statistics (mean, median, mode, range, standard deviation) and inferential statistics (sign test, Wilcoxon, Mann-Whitney, Spearman's rho, Pearson's r, related and unrelated t-tests, Chi-square).
• →Reliability and validity: internal and external validity, test-retest and inter-rater reliability, and threats such as demand characteristics, investigator effects, and social desirability bias.

• Identifies manipulation of the independent variable by the researcher as the defining feature of an experiment
• States that the dependent variable is measured to detect the effect of the manipulation
• Explains that control of extraneous variables allows a cause-and-effect conclusion to be drawn
• Distinguishes a true experiment from a quasi-experiment by noting random allocation of participants to conditions
• Evaluates experiments by weighing high control and replicability against artificiality and reduced ecological validity
• Identifies that the independent variable in a quasi-experiment is an existing participant characteristic rather than a manipulated one
• States that participants are not randomly allocated to conditions in a quasi-experiment
• Explains that the absence of random allocation weakens causal inference because confounding variables may differ between groups
• Distinguishes a quasi-experiment from a true experiment by focusing on how groups are formed
• Evaluates quasi-experiments by weighing real-world applicability against reduced control over extraneous variables
• Defines participant observation as a method where the observer joins and takes part in the group or setting being studied.
• Identifies participant observation correctly when a scenario describes a researcher joining a group and engaging in its activities.
• Explains a strength such as gaining deeper insight or access to behaviour that outsiders would not normally see.
• Explains a weakness such as reduced objectivity, observer bias or the risk of becoming too involved with the group.
• Recognises an ethical issue such as difficulty obtaining informed consent or protecting participants' privacy.
• Defines non-participant observation as a method where the observer watches and records behaviour without joining the group's activities.
• Identifies non-participant observation correctly when a scenario describes a researcher observing from outside the group.
• Explains a strength such as maintaining detachment or making objective recording easier.
• Explains a weakness such as reduced insight into participants' meanings or behaviour changing because they are watched.
• Recognises a practical or ethical issue such as difficulty gaining access or problems with consent and privacy.
• Defines content analysis as a method that examines existing material such as text, transcripts or media and sorts recurring features into categories
• Identifies that the researcher builds a coding system with defined categories before counting or describing occurrences
• Explains that content analysis can be quantitative through frequency counts or qualitative through thematic description
• States that inter-coder or inter-rater reliability is checked by having more than one coder apply the same categories
• Recognises that content analysis uses material that already exists, which raises fewer ethical issues than generating new data
• Notes that researcher-defined categories can introduce subjectivity as a limitation
• Defines structured interviews and questionnaires as self-report methods using the same pre-set questions in the same order for every participant
• Distinguishes a structured interview, delivered face to face by the researcher, from a questionnaire, usually completed in writing by the participant
• Explains that fixed questions and closed response options such as Likert scales make data easy to compare and analyse, raising reliability
• Identifies that closed questions can force participants into categories and that participants may give socially desirable answers
• Notes that open questions can be added to gain richer qualitative data
• States that standardised instructions, piloting and ethical safeguards such as informed consent and confidentiality are required
• Identifies a semi-structured interview as one using a pre-planned set of core questions or topic areas combined with unscripted follow-up probes.
• Distinguishes semi-structured from structured interviews by noting that structured interviews keep every question and wording fixed for all participants.
• Distinguishes semi-structured from unstructured interviews by noting that unstructured interviews begin with little or no predetermined schedule.
• Explains that the prepared schedule gives comparable coverage of topics while the probes allow depth and clarification.
• Evaluates the design by weighing rich, flexible qualitative data against interviewer effects, time cost and reduced standardisation.
• Defines a correlational study as one measuring two co-variables in the same participants without manipulating either variable.
• States that the correlation coefficient ranges from −1 to +1, with values near +1 indicating a strong positive relationship and near −1 a strong negative relationship.
• Explains that a coefficient near 0 indicates little or no linear relationship between the co-variables.
• Explains that correlation does not demonstrate causation because a third variable may affect both co-variables or the direction of influence may be reversed.
• Evaluates correlational studies by noting they suit variables that cannot ethically be manipulated and permit prediction, but cannot establish cause and effect.
• Identifies an in-depth study of a single individual, small group or single event as the defining feature of a case study
• Recognises that case studies typically combine several methods and data sources rather than relying on one technique
• Selects the option stating that findings from a single case are difficult to generalise to the wider population
• Distinguishes a case study from a laboratory experiment by noting the absence of tight manipulation and control of variables
• Acknowledges that case studies are often used where experimental study would be unethical or impossible
• Identifies fMRI and PET as functional techniques that measure brain activity rather than structure alone
• Recognises MRI and CT as structural techniques that image brain anatomy
• Selects the option stating that scans produce objective, quantifiable data that can be compared between participants
• Notes that scanning data are correlational, so a link between activity and behaviour does not establish cause
• Acknowledges practical limitations such as cost, specialist equipment and restricted movement during scanning
• Identifies that the same participants are followed and measured repeatedly over an extended period.
• Explains that within-participant comparison controls individual differences when assessing change.
• Recognises attrition, repeated-testing effects and high cost as limitations of the design.
• Distinguishes longitudinal from cross-sectional designs by the use of repeated measures on one sample.
• Identifies that different groups of participants are compared at a single point in time.
• Explains that each participant is measured once, so no within-person change is tracked.
• Recognises cohort and individual differences as threats to interpreting group differences as change.
• Contrasts the design with longitudinal studies, which repeatedly measure the same sample.
• Identifies a self-report when the participant supplies information about their own experiences, opinions or behaviour.
• Distinguishes self-report methods such as questionnaires and interviews from observation, where the researcher records behaviour.
• Recognises that self-report data may be quantitative, qualitative or both, depending on the questions asked.
• Explains that self-reports can be affected by social desirability bias and demand characteristics.
• States that self-reports rely on participants' own accounts rather than direct measurement of behaviour.
• Identifies quantitative data as numerical information such as test scores or closed-scale ratings.
• Identifies qualitative data as non-numerical information such as interview transcripts or open-ended responses.
• Recognises that a report meeting the requirement includes both numerical and non-numerical data.
• Explains that combining both data types allows statistical comparison alongside richer description of experience.
• Distinguishes a mixed-methods design from one that collects only one data type.
• Identifies a primary source as data collected first-hand by the researcher for the current aim
• Identifies a secondary source as data already gathered by others and analysed by the researcher
• Explains that including both types allows first-hand data to be combined with existing evidence
• Evaluates a strength or limitation of using primary sources, such as control over collection versus time cost
• Evaluates a strength or limitation of using secondary sources, such as availability versus lack of control over how data were produced

• 💡Name the IV and DV explicitly in any scenario before deciding whether it is an experiment
• 💡Use the phrase 'manipulation of the IV' to justify your classification rather than relying on the setting
• 💡When evaluating, link each strength or weakness to a named feature such as control or ecological validity
• 💡Identify the grouping variable and confirm it is an existing characteristic, not something the researcher manipulated
• 💡Use the phrase 'no random allocation' to justify classifying a study as a quasi-experiment
• 💡When evaluating, link reduced control to specific confounding participant variables that could explain results
• 💡Read the scenario carefully and decide whether the observer joins the group before choosing participant observation.
• 💡Link each strength or weakness to the specific study described rather than listing generic points.
• 💡Use precise terminology such as observer bias, ecological validity and informed consent to gain credit.
• 💡Check whether the observer takes part in activities before selecting non-participant observation.
• 💡Tie each strength or weakness to the scenario, for example reactivity when participants know they are watched.
• 💡Use precise terms such as detachment, reactivity and ecological validity to strengthen your answer.
• 💡Name the coding system and categories explicitly when describing the method, because examiners look for procedural detail
• 💡Give a concrete example such as counting aggressive acts in children's television programmes to show how categories become counts
• 💡When evaluating, balance a strength such as high ecological validity against a limitation such as researcher subjectivity
• 💡Use the term standardised when describing the fixed questions and order, because it signals reliability
• 💡Give a concrete example such as a Likert-scale item measuring stress to show how closed questions produce comparable data
• 💡When evaluating, weigh reliability from standardisation against validity problems from forced-choice items and social desirability
• 💡Anchor your answer on the defining feature: a prepared schedule plus flexible probes, and state both parts explicitly.
• 💡When asked to evaluate, give one strength and one limitation and link each to the purpose of the study rather than listing generic points.
• 💡Use the term 'standardisation' precisely: explain that semi-structured interviews are less standardised than structured ones because probes vary.
• 💡Always name both co-variables and state that neither is manipulated when identifying a correlational design.
• 💡When interpreting a coefficient, report both its strength and its direction, for example a strong negative correlation.
• 💡Link the 'no causation' limitation to a concrete alternative explanation such as a third variable rather than stating it as a slogan.
• 💡Underline the key term in the question stem before reading the options, so you know whether it asks for a definition, a strength or a limitation
• 💡Eliminate options that describe a different method, such as an experiment with a control group, before choosing between the remaining two
• 💡Check whether the option refers to depth of detail or to breadth of sample, because case studies are characterised by depth
• 💡Sort the options into structural and functional techniques before deciding which one answers the question
• 💡Look for the word causation in an option and treat it with caution, because scan data are correlational
• 💡Check whether the question asks about a strength or a limitation, since the same feature can be framed either way
• 💡Underline the phrase 'same participants over time' in the stem before choosing an option.
• 💡Eliminate any option that describes independent groups or a single testing occasion.
• 💡Link the chosen option to one strength or limitation, such as control of individual differences or attrition.
• 💡Look for 'different groups' and 'one time point' as the signature of a cross-sectional design.
• 💡Reject options that mention following the same people over years.
• 💡State one advantage, such as speed and no attrition, when justifying your choice.
• 💡Read the procedure carefully and ask who provides the information before choosing an answer.
• 💡Eliminate options that describe the researcher recording behaviour directly, because those describe observation.
• 💡Link each option to a named feature such as demand characteristics or closed questions rather than judging by familiarity.
• 💡Check whether the option confuses the method with the design, since a self-report can be used within a correlation or a quasi-experiment.
• 💡Classify each piece of evidence in the scenario as numerical or non-numerical before choosing.
• 💡Reject options that describe only one data type when the statement requires both.
• 💡Look for words such as open questions, transcripts or quotations to confirm qualitative data.
• 💡Look for scores, frequencies or ratings to confirm quantitative data.
• 💡Read the scenario carefully and name the specific primary and secondary source before evaluating them
• 💡Link each evaluation point to the stated aim of the study rather than giving generic strengths and limitations
• 💡Use precise terms such as first-hand collection and existing data to show understanding
• 💡Always use precise methodological terminology (e.g., 'operationalise', 'counterbalance', 'inter-rater reliability') rather than everyday language. This demonstrates secure subject knowledge and accesses higher mark bands.
• 💡When evaluating, use a 'point-evidence-explain' structure: make a point, support it with evidence from the stimulus or your own knowledge, and explain why it is a strength or weakness in that specific context.
• 💡For statistical tests, always state the observed value, the critical value, the degrees of freedom or N, the significance level, and whether the result is significant. Then relate this back to the experimental hypothesis.

• Confuses correlation with causation; state that only a manipulated IV with controlled extraneous variables supports a causal claim
• Treats any questionnaire study as an experiment; check whether the researcher actively manipulated an IV and allocated participants
• Assumes all experiments occur in a laboratory; recognise field and natural experiments as variants with different control levels
• Claims that quasi-experiments randomly allocate participants; state that groups are formed by an existing characteristic
• Treats a quasi-experiment as identical to a true experiment; highlight the missing random allocation as the key difference
• Assumes quasi-experiments cannot be useful; recognise their value for studying naturally occurring groups despite weaker control
• Confuses participant observation with a non-participant role; keep the observer inside the group and taking part.
• Treats participant observation as always covert; note that it can be overt or covert and the scenario decides which.
• Claims participant observation removes all researcher influence; accept that the observer's presence can still change behaviour.
• Assumes participant observation guarantees objectivity; recognise that involvement can increase bias.
• Confuses non-participant observation with joining the group; keep the observer separate from the activities.
• Assumes the observer's separation removes all influence on behaviour; accept that being watched can still cause reactivity.
• Treats non-participant observation as always overt; note that it can be overt or covert and the scenario decides which.
• Claims non-participant observation gives full insight into participants' meanings; recognise that an outsider may miss these.
• Confuses content analysis with thematic analysis of interview data the researcher has just collected; keep content analysis focused on analysing existing material through a coding system
• Assumes content analysis is always quantitative; recognise that it can also be qualitative and descriptive
• Ignores reliability checks; remember that two coders should apply the same categories to test consistency
• Treats content analysis as a form of experiment; describe it as an observational, non-experimental method
• Confuses structured interviews with unstructured interviews; keep the focus on identical pre-set questions in a fixed order
• Assumes questionnaires are always completed face to face; recognise that they are usually self-administered in writing or online
• Overlooks social desirability bias; remember that participants may answer in ways they think are acceptable
• Ignores the need for piloting; describe how a pilot study checks that questions are clear before the main study
• Treats a semi-structured interview as having no prepared questions at all; correct this by stating that a core schedule of questions or topics is always prepared in advance.
• Confuses semi-structured with structured interviews by claiming every question and probe is fixed; correct this by noting that follow-up probes are unscripted and vary between participants.
• Assumes semi-structured interviews produce numerical scores by default; correct this by recognising that they typically generate qualitative data, though responses can be coded or themed.
• Ignores interviewer effects; correct this by acknowledging that the interviewer's tone, prompts and presence can shape what participants say.
• Claims a strong correlation proves that one variable causes the other; correct this by stating that correlation shows co-variation only and causation requires further evidence.
• Describes the researcher as manipulating an independent variable; correct this by noting that a correlational study measures co-variables as they naturally occur.
• Treats a coefficient of −0.8 as weaker than +0.4 because of the minus sign; correct this by explaining that −0.8 indicates a stronger relationship, with the sign showing direction.
• States that a coefficient of 0 means the variables are unrelated in every respect; correct this by saying it indicates no linear relationship, as a non-linear pattern may still exist.
• Confuses a case study with any small-sample experiment; correct this by noting that a case study explores a case in depth without necessarily manipulating an independent variable
• Assumes a case study uses only one method; correct this by recognising that case studies commonly triangulate interviews, observation and records
• Claims case study findings generalise easily to all people; correct this by stating that a sample of one limits generalisability
• Treats a large-scale questionnaire survey as a case study; correct this by checking whether the focus is one detailed case rather than many participants
• Treats every brain scan as a measure of activity; correct this by separating structural techniques such as MRI from functional techniques such as fMRI
• Assumes a scan showing activity in a region proves that region causes the behaviour; correct this by describing the finding as correlational
• Ignores practical constraints and presents scanning as cheap and easy; correct this by noting cost, equipment and the need to remain still
• Confuses the measure used by a technique, such as blood oxygenation for fMRI, with an unrelated measure; correct this by matching each technique to its actual index
• Confuses longitudinal with cross-sectional by describing several age groups tested once; correct this by stating that one sample is measured at successive time points.
• Treats a single follow-up test as longitudinal; correct this by requiring at least two measurement waves over an extended period.
• Ignores sample attrition across waves; correct this by noting that drop-out can bias later comparisons.
• Describes the same participants tested repeatedly; correct this by stating that separate groups are compared once.
• Claims the design directly measures change within individuals; correct this by noting it compares different people at one moment.
• Overlooks cohort effects; correct this by explaining that group differences may not reflect genuine developmental change.
• Labels any study using a questionnaire as an experiment; check whether an independent variable is manipulated before naming the design.
• Assumes self-reports always produce qualitative data; recognise that closed rating scales yield numerical scores.
• Treats observation of participants as a self-report; identify who records the information before deciding.
• Ignores social desirability bias; consider that participants may present themselves favourably.
• Treats a closed questionnaire alone as satisfying the requirement; add open-ended items or interviews to obtain qualitative data.
• Assumes qualitative data are inferior and can be omitted; recognise that the statement requires both types.
• Confuses the data type with the method name; judge by whether the information is numerical or non-numerical.
• Describes a study collecting only numbers as mixed-methods; check that words as well as numbers are gathered.
• Labels any published material as primary because it appears in a textbook; check whether the researcher collected it first-hand for the current aim
• Treats a secondary source as automatically unreliable; weigh its strengths and limitations rather than dismissing it
• Assumes both sources must be the same type of method; recognise that primary and secondary refer to the origin of the data, not the technique
• Students often think that a reliable study is automatically valid. Correction: Reliability is necessary but not sufficient for validity; a study can be consistent but measure the wrong thing.
• Students frequently confuse random sampling with random allocation. Correction: Random sampling refers to how participants are selected from a population; random allocation refers to how participants are assigned to conditions in an experiment.
• Students often believe that a p-value of 0.05 means there is a 5% chance the results are due to chance. Correction: It means that if the null hypothesis is true, there is a 5% probability of obtaining the observed results (or more extreme) by chance.

1. 1Days 1-2: Create a glossary of key methodological terms (IV, DV, operationalisation, reliability, validity, sampling, designs). Use flashcards to test yourself.
2. 2Days 3-4: For each experimental design and sampling method, write one strength and one weakness, and apply them to a novel scenario. Practice identifying IVs and DVs in past paper questions.
3. 3Days 5-6: Learn the criteria for choosing inferential statistical tests (type of data, design, test of difference or association). Create a decision tree and practice with past data sets.
4. 4Days 7-8: Practice writing 6-mark evaluation answers using the 'point-evidence-explain' structure. Focus on reliability, validity, and ethics in context.
5. 5Days 9-10: Complete a full past paper under timed conditions, then mark it using the mark scheme. Review any misconceptions and revisit weak areas.

• 📋Short-answer questions (1-3 marks) asking students to identify variables, sampling methods, or ethical issues from a scenario. Advice: read the scenario carefully and use the exact terms from the specification.
• 📋Structured questions (4-6 marks) requiring explanation of how a methodological issue could be addressed or evaluated. Advice: use the context and link each point to the scenario.
• 📋Data analysis questions (4-8 marks) involving calculation of descriptive statistics or selection and interpretation of inferential tests. Advice: show your working and always state the conclusion in terms of the hypothesis.
• 📋Extended response questions (10-15 marks) requiring evaluation of methodological issues in a given study. Advice: plan your answer, use a range of evaluation points, and refer to the stimulus throughout.