Skip to main content

Validity — Previous Year Questions

📊 Topic Overview​

MetricValue
Papers Appeared16 of 28 (57.1%)
Priority🔴 HIGH — dominates December papers
Section DistributionSec A: 5
Most RepeatedThreats to internal validity (11 times)
Session PatternHeavily December
ProbabilityExtremely High in December

🎯 Must-Know Concepts​

  1. Definition of validity and types (internal, external, construct, statistical conclusion)
  2. Threats to internal validity (Campbell and Stanley — 8 threats)
  3. Threats to external validity
  4. Distinction between internal and external validity

🔥 Most Repeated Question: Types of Validity and Threats to Internal Validity (11 appearances)​

Answer (Section A — ~450 words):​

Validity refers to the degree to which a research study measures what it claims to measure or, in the context of experimental design, whether the conclusions drawn from a study are justified. Cook and Campbell identified four types of validity relevant to research design.

Types of Validity:

Internal Validity: The degree to which changes in the dependent variable can be attributed to the independent variable rather than to confounding factors. Internal validity is the sine qua non of experimental research — without it, causal inferences are unwarranted.

External Validity: The degree to which findings can be generalised to other people, settings, times, and operationalisations. A study can be internally valid but not generalisable if it used a highly unrepresentative sample or artificial setting.

Construct Validity: The degree to which the operationalisation of variables accurately represents the theoretical constructs they are intended to measure. Poor construct validity means the study is testing something different from what it claims.

Statistical Conclusion Validity: The degree to which the statistical procedures used are appropriate and the conclusions about covariation are correct. Violations include low statistical power, violated statistical assumptions, and fishing for significance.

Threats to Internal Validity (Campbell and Stanley, 1963):

  1. History: Extraneous events occurring between pre-test and post-test that affect the DV. Controlled by a comparison group.
  2. Maturation: Participants change over time (grow older, become fatigued) independently of the IV. Controlled by time-limited studies or matched controls.
  3. Testing: Exposure to a pre-test sensitises participants and affects post-test scores. Controlled by a Solomon four-group design.
  4. Instrumentation: Changes in measuring instruments or observers over time. Controlled by standardised measurement.
  5. Statistical Regression (Regression to the Mean): Groups selected on extreme scores tend to score closer to the mean on retesting regardless of treatment. Controlled by not selecting extreme groups or using randomisation.
  6. Selection Bias: Pre-existing differences between groups due to non-random assignment. Controlled by random assignment.
  7. Experimental Mortality (Attrition): Differential dropout of participants from groups. Controlled by examining who drops out and why.
  8. Selection-Maturation Interaction: Different groups mature at different rates. Particularly problematic in quasi-experimental designs.

Threats to External Validity: Population validity: Whether the sample represents the population. Ecological validity: Whether the laboratory setting represents natural settings. Temporal validity: Whether findings hold across different time periods. Reactive effects: Participants behave differently knowing they are being studied (Hawthorne effect).

In conclusion, internal validity is essential for causal inference; external validity determines the scope of generalisation. Both must be considered in research design, though they often trade off — tightly controlled laboratory studies maximise internal validity at the cost of external validity.

📝 Other Questions​

Differentiate internal and external validity (Section B — ~250 words)​

DimensionInternal ValidityExternal Validity
DefinitionCausal relationship between IV and DV is unambiguousFindings generalise beyond the study context
Question"Did X cause Y in this study?""Do these findings apply elsewhere?"
ThreatsHistory, maturation, testing, instrumentation, regression, selection, mortalitySampling bias, reactive effects, setting specificity
Enhanced byRandom assignment, control groups, standardisationRandom sampling, multiple settings, replication
Trade-offHigh internal validity often requires artificial conditionsNaturalistic settings reduce control

A study with high internal validity but low external validity establishes a causal relationship in a narrow, artificial context that may not generalise. Ideally both are high, but researchers must be explicit about which is prioritised and why.

🧠 Memory Aids​

8 Threats to Internal Validity (HMTISSEM): History, Maturation, Testing, Instrumentation, Statistical regression, Selection, Experimental mortality, Maturation-selection interaction

📚 Related Content​

✅ Self-Assessment​

  1. List the 8 threats to internal validity identified by Campbell and Stanley.
  2. How does statistical regression threaten internal validity?
  3. What is the Hawthorne effect and which type of validity does it threaten?
  4. Why does random assignment control for selection bias?
  5. How does the Solomon four-group design control for the testing threat?

💡 Exam Tips​

  • Section B (dominant): Types of validity (4) → Threats to internal validity (list and explain each) → Brief mention of external validity threats.
  • All 8 internal validity threats must be named — use the mnemonic. Partial credit is given for each.
  • December papers almost always include a validity question — this is the highest-priority December topic.