tests and intelligence

1. Standardized Test Construction

  • Designed by psychometricians

  • Steps:

    • Write questions that differentiate ability levels

    • Pilot testing → test on representative sample to find flaws

    • Item analysis → remove questions too easy/hard

    • Establish norms (average performance baseline)

  • Goal: create a test that is reliable + valid

2. Norms & Scoring

  • Norm-referenced tests compare individuals to a typical group

  • Mean score used to define:

    • Passing vs failing

  • Based on normal distribution

3. Reliability (Consistency)

  • Test gives stable, repeatable results

  • Types:

    • Test-retest → same test over time

    • Split-half → compare halves of test

    • Equivalent forms → different versions, same results

4. Validity (Accuracy)

  • Measures what it is supposed to measure

  • Types:

    • Content validity → covers full material

    • Criterion validity → matches external measure

      • Predictive → future performance (SAT → college)

      • Concurrent → current performance

    • Construct validity → measures intended concept (true validity)

5. Intelligence (Definition)

  • Ability to:

    • Learn

    • Solve problems

    • Adapt to new situations

6. Theories of Intelligence

  • Spearman

    • “g factor” (general intelligence)

    • “s factors” (specific abilities)

  • Sternberg (Triarchic Theory)

    • Analytical (academic)

    • Creative (innovation)

    • Practical (real-world skills)

  • Gardner (Multiple Intelligences)

    • 8 types (linguistic, logical, musical, spatial, etc.)

    • Challenges single IQ idea

7. IQ & Measurement

  • IQ = (Mental Age / Chronological Age) × 100

  • Mean = 100, SD = 15

  • Scores follow normal distribution

  • Percentiles show relative standing

8. Testing Issues & Bias

  • Test anxiety affects performance

  • Examiner familiarity improves scores

  • Bias factors:

    • Race

    • Gender

    • Language

  • Environmental + genetic factors both influence IQ

Core Takeaway

A good psychological test must be:

  • Standardized

  • Reliable (consistent)

  • Valid (accurate)

  • Fair (unbiased and representative)