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

Validity is a cornerstone concept in psychological research, referring to the accuracy of a measure or study in assessing what it intends to assess. In the Eduqas A-Level specification, students must understand and apply different types of validity: internal (cause-effect), external (generalisation), construct (operationalisation), face (superficial appearance), concurrent (correlation with existing measures), and predictive (forecasting future outcomes). This topic is essential for evaluating research methods, designs, and data collection techniques across all areas of psychology.

Understanding validity allows students to critically assess the strengths and limitations of studies, from classic experiments like Milgram's obedience research to contemporary questionnaire-based studies. It connects closely with reliability, though the two are distinct: a measure can be reliable but not valid. Mastery of validity is crucial for high-scoring evaluation answers in exams, as it demonstrates analytical depth and the ability to apply methodological concepts to specific contexts.

• →Internal validity: the extent to which the IV genuinely causes the DV, free from confounding variables and extraneous influences.
• →External validity: the extent to which findings can be generalised to other settings (ecological), populations (population validity), and times (temporal validity).
• →Construct validity: the extent to which operational definitions (e.g., how 'aggression' is measured) accurately reflect the theoretical construct.
• →Face validity: a subjective, superficial judgement of whether a measure appears to test what it claims to test.
• →Concurrent and predictive validity: concurrent validity is established by correlating a new measure with an existing valid measure; predictive validity is established by correlating scores with future behaviour or outcomes.

• Defines internal validity as the extent to which a study measures what it intends and shows a causal effect of the independent variable on the dependent variable.
• Identifies a specific threat such as demand characteristics, investigator effects, order effects or uncontrolled participant variables in the described study.
• Explains how the identified threat could produce the observed result instead of the manipulation, giving a causal mechanism.
• Judges whether the design controls the threat, for example through random allocation, standardised procedures or counterbalancing.
• Defines external validity as the extent to which findings can be generalised beyond the sample, setting, task or time of the study.
• Distinguishes forms of external validity, naming population, ecological and temporal validity as relevant to the described study.
• Explains why a specific feature, such as a narrow sample or artificial task, limits generalisation to a stated wider group or setting.
• Judges how far the findings transfer, weighing the limitation against features that support generalisation.
• Identifies researcher bias as the investigator's expectations influencing design, procedure, recording or interpretation of results.
• Explains demand characteristics as cues in the study that communicate the expected behaviour to participants, who then adjust their responses.
• Describes social desirability as participants presenting themselves favourably, giving socially acceptable rather than truthful answers.
• Links each threat to reduced validity by showing how it introduces a systematic influence separate from the variable being studied.
• Applies a named threat to a concrete research scenario, such as a self-report questionnaire on sensitive behaviour producing socially desirable answers.
• Proposes standardised procedures, clear operational definitions or independent second raters to reduce researcher bias.
• Suggests single-blind procedures, neutral standardised instructions or concealment of the hypothesis to reduce demand characteristics.
• Recommends anonymity, confidentiality or less threatening question formats to reduce social desirability.
• Explains how the chosen control reduces the specific threat and thereby improves validity.
• Applies an appropriate control to a described study, matching the remedy to the threat present.
• Identifies concurrent validity as agreement between a new measure and an established measure administered at approximately the same time
• Identifies predictive validity as the correlation between a test score and a criterion measured at a later date
• Describes face validity as a superficial judgement that items appear relevant to the construct
• Explains content validity as the extent to which items adequately sample the full domain
• Explains construct validity as the degree to which a test measures the theoretical construct it claims to measure
• Selects the validity type that matches the evidence described in the stem rather than the one that merely sounds most rigorous

• 💡Read the stem for the design features first, then name the single most serious threat before judging the study.
• 💡Link each threat to the specific dependent variable measured, not to validity in general.
• 💡Use the phrase 'the change in the dependent variable may be caused by' to keep the causal focus.
• 💡Name the form of external validity at stake before explaining the limitation.
• 💡State the wider population, setting or time to which generalisation is claimed, then judge the fit.
• 💡Balance the limitation with any feature that supports generalisation to reach a reasoned judgement.
• 💡Learn each threat with a one-line definition and a short example so you can recognise it in an unfamiliar scenario.
• 💡When a question gives a study, name the specific threat and point to the feature of the design or procedure that produces it.
• 💡Distinguish the three threats clearly, since multiple-choice distractors often swap their definitions.
• 💡Build a mental table pairing each threat with at least one practical control so you can select the best match quickly.
• 💡When applying to a scenario, name the threat, propose the control and state the improvement to validity in one linked chain.
• 💡Watch for distractors that describe a control for a different threat than the one in the question.
• 💡Underline the timing words in the stem, such as same time or later, because they separate concurrent from predictive validity
• 💡Match the named validity type to the specific evidence offered rather than choosing the most impressive-sounding term
• 💡Revise construct validity as the umbrella concept and link it to convergent and discriminant evidence
• 💡Always specify the type of validity you are discussing (e.g., 'internal validity', 'ecological validity') and link it explicitly to the study or measure in the question.
• 💡When evaluating, use the 'threat -> effect -> improvement' structure: identify the threat, explain how it reduces validity, and suggest a concrete way to enhance validity.
• 💡Use named studies and specific methodological terms (e.g., 'demand characteristics', 'social desirability bias', 'counterbalancing') to demonstrate precise knowledge and application.

• Confuses internal validity with reliability; keep internal validity about causal control and measurement of the intended variable.
• Treats any laboratory setting as automatically high in internal validity; check whether confounds and demand characteristics are actually controlled.
• Lists a threat without explaining how it changes the dependent variable; state the mechanism linking the threat to the result.
• Treats external validity as a single idea; separate population, ecological and temporal validity when judging a study.
• Assumes a laboratory study always lacks ecological validity; judge the specific task and setting against real-world behaviour.
• Confuses external validity with internal validity; keep external validity about generalisation rather than causal control.
• Confuses demand characteristics with social desirability; keep them distinct by treating demand characteristics as cues about the study's hypothesis and social desirability as a general wish to look good.
• Treats researcher bias as random error; recognise it as a systematic influence that pushes results in a particular direction.
• Describes a threat without linking it to validity; state explicitly how the influence undermines the claim that the study measures what it intends.
• Offers a generic control without naming the threat it addresses; state which validity issue the measure targets and how it works.
• Confuses reliability measures with validity controls; explain how a control reduces systematic bias rather than merely improving consistency.
• Suggests a control that creates a new problem, such as deceiving participants without debriefing; check that the remedy is ethical and does not introduce fresh bias.
• Confuses concurrent validity with predictive validity; check whether the criterion is measured at the same time or later
• Treats face validity as strong evidence of a good test; recognise that it is only a surface judgement
• Assumes content validity is established by the researcher alone; note that expert or independent review of domain coverage is typical
• Equates reliability with validity; remember that a measure can be consistent yet still measure the wrong thing
• Students often think validity and reliability are the same. Correction: Reliability is about consistency; validity is about accuracy. A scale can consistently give the same weight (reliable) but be wrongly calibrated (invalid).
• Students may believe that a lab experiment is always high in validity. Correction: Lab experiments often have high internal validity but low ecological validity due to artificial settings and tasks.
• Students sometimes assume that a large sample size guarantees validity. Correction: A large sample improves population validity but does not address internal validity threats such as demand characteristics or confounding variables.

1. 1Day 1-2: Learn definitions and examples of each type of validity (internal, external, construct, face, concurrent, predictive). Create flashcards with definitions and a study example for each.
2. 2Day 3-4: Practice applying validity concepts to at least three different studies from your course (e.g., a lab experiment, a questionnaire study, a naturalistic observation). For each, identify threats and suggest improvements.
3. 3Day 5-6: Complete past paper questions on validity, focusing on 4-6 mark structured questions. Mark your answers using the mark scheme and note where you missed specific terminology.
4. 4Day 7-8: Review common misconceptions and examiner tips. Write a model answer for a 6-mark question on internal validity and another on external validity.
5. 5Day 9-10: Self-test using active recall prompts and teach the concept to a peer. Consolidate by creating a mind map linking validity to reliability, sampling, and research design.

• 📋Definition and explanation questions: e.g., 'Explain what is meant by internal validity' (2-4 marks). Advice: give a precise definition and use an example to illustrate.
• 📋Application questions: e.g., 'Identify one threat to validity in this study and explain how it could be reduced' (4-6 marks). Advice: link the threat to the specific study context and suggest a practical control.
• 📋Evaluation questions: e.g., 'Discuss the validity of the study you described in part (a)' (6-8 marks). Advice: cover multiple types of validity, use the 'threat-effect-improvement' structure, and refer to named studies.
• 📋Comparison questions: e.g., 'Explain the difference between concurrent validity and predictive validity' (4 marks). Advice: define both terms clearly and give an example of how each is assessed.