Validity

• Validity refers to whether a test, study, or measurement actually measures what it claims to measure.

  • If a study is valid, the conclusions drawn are accurate and meaningful.
• Internal Validity is whether the results of a study are due to the manipulation of the independent variable and not other factors (extraneous/confounding variables). It applies to experiments (especially lab studies).

  • Threats:
  • Demand characteristics
  • Investigator effects
  • Confounding variables
• External Validity is how well the findings generalise beyond the research setting to other settings (ecological validity), other populations (population validity) and other eras (temporal validity).
• Ecological Validity is a type of external validity.

  • Can results be generalised to other settings, real-world settings in particular
  • Low in artificial settings (e.g., lab memory tests)
• Population Validity is a type of external validity.

  • Can results be generalised to other people?
  • Is the sample representative of the wider population?
• Temporal Validity is a type of external validity.

  • Can results be generalised across time periods?
  • Is the research still relevant today or outdated?
• Face Validity is whether a test appears to measure what it’s supposed to.

  • Example: A maths test with only maths problems = high face validity.
• Concurrent Validity is how well a new test compares to an existing valid test. This is demonstrated when the results obtained are very close to or match those obtained on the existing valid test.

  • How to assess: Give participants both tests → calculate a correlation between the results.
  • Example: A new anxiety scale is tested against the Beck Anxiety Inventory. If correlation is high (e.g. r ≥ +0.80) → high concurrent validity.
• Improving validity in experiments:

  • Use control groups: assess whether changes in the dependent variable were due to the effect of the independent variable
  • Use single-blind or double-blind procedures: reduces demand characteristics and investigator effects
  • Standardise procedures: reduces effects of situational variables
• Improving validity in questionnaires:

  • Assure anonymity: participants are less likely to give inaccurate data if they know their data will remain anonymous
  • Avoid leading questions: ensures answers are accurate and honest
• Improving validity in observations:

  • Use covert methods: behaviour is likely to be natural and authentic
  • Clear behavioural categories: broad, overlapping or ambiguous categories may have a negative impact on validity
• Improving validity in qualitative research:

  • Triangulation: using multiple sources of evidence
  • Participant validation: checking interpretations with participants to ensure accuracy