PAR Lect 5 Validity

18 Questions

Settings
Please wait...
PAR Lect 5 Validity

PAR Lect 5 VALIDITY


Questions and Answers
  • 1. 
    All different components of reliability assess different forms of error. This is:
    • A. 

      True

    • B. 

      False

  • 2. 
    The view of face validity ie., does it measure what it's intended to measure, is largely a reaction to the:
    • A. 

      WAIS

    • B. 

      WISC

    • C. 

      WIAT

    • D. 

      MMPI-2

  • 3. 
    We don't talk about a test being valid as such, but only in terms of the ___ we are drawing from the test.
    • A. 

      Inference

    • B. 

      Score

    • C. 

      Results

    • D. 

      Effect

  • 4. 
    Validity is:
    • A. 

      The inference from the data being produced by the test.

    • B. 

      Requires ongoing empirical investigation.

    • C. 

      Is a matter of degree.

    • D. 

      Attaches to the inferences drawn from the test scores.

    • E. 

      ALL OF THE ABOVE

  • 5. 
    We don't talk about a test being valid because the V of a test is for on going empirical investigation, the evidence is accumulated overtime and so the V evidence will change from 1 period to another but also from 1 group to another. This is:
    • A. 

      TRUE

    • B. 

      FALSE

  • 6. 
    There are 3 classic types of measurement validity:
    • A. 

      Construct

    • B. 

      Content

    • C. 

      Criterion

    • D. 

      Conversion

    • E. 

      A-c

  • 7. 
    Content-related evidence relates to:
    • A. 

      The V of the item sample selection

    • B. 

      Face validity

    • C. 

      Concurrent V

    • D. 

      Predictive V

    • E. 

      A and b

  • 8. 
    In content V evidence we are talking about:
    • A. 

      How well the scale checked against the relevant content domain.

    • B. 

      Academic and edu'al tests of attainment and ability.

    • C. 

      Is it representative of the population?

    • D. 

      It's a judgement call.

    • E. 

      All of the above

  • 9. 
    Face validity evidence:
    • A. 

      The scale content seems valid "on its face".

    • B. 

      So for eg you know the DAS is about road rage.

    • C. 

      A judgement call

    • D. 

      All of the above

  • 10. 
    Criterion validity (CV) of a scale is about the degree to which the scores on the scale:
    • A. 

      Agree or predict or correlate with some empirical criteria.

    • B. 

      Concurrent CV demos when scores corr with est measures simultan.

    • C. 

      So did the DAS corr betw hostile and non-hostile drivers?

    • D. 

      Or existing anger measures, eg., spielberger trait anger scale?

    • E. 

      All of the above

  • 11. 
    Predictive CV evidence is the ability of the measure:
    • A. 

      To pred something it should be able to predict in the future.

    • B. 

      A corr'n is used as an index of temporal relationships.

    • C. 

      We could admin the DAS pre-licence and observe the future.

    • D. 

      All of the above

  • 12. 
    Construct-related validity is supported by the gradual accumulation of 2 sorts of evidence:
    • A. 

      Convergent and discriminant evidence.

    • B. 

      Convergent is the extent the items converge on the construct.

    • C. 

      So diffenbacher only wants to measure the driving anger construct

    • D. 

      NOT Type a personality, extroversion, self-esteem or egotism etc.

    • E. 

      To what extent is he examining driving anger and nothing else?

    • F. 

      So to what extent do the items converge on 'driving anger'?

    • G. 

      All of the above

  • 13. 
    So, convergent V says:
    • A. 

      Measures that are related should be strongly correlated.

    • B. 

      And discriminant V says: measures of different constructs

    • C. 

      Should not correlate highly with each other.

    • D. 

      All of the above

  • 14. 
    A more sophisticated and complex way of determining the construct validity of a scale is using:
    • A. 

      Inter-item correlations

    • B. 

      Multitrait-multimethod matrix

    • C. 

      Multiple scales

    • D. 

      None of the above

  • 15. 
    The multitrait multimethod matrix is a matrix of correlations that simultaneously addresses both __ and __ validities, at the same time.
    • A. 

      Construct, criterion

    • B. 

      Content, criterion

    • C. 

      Concurrent, convergent

    • D. 

      Convergent, discriminant

  • 16. 
    The multitrait multimethod matrix addresses the importance of measurement method OR method error as a possible confound or METHOD BIAS. It also has some indication of reliability but not validity. This is:
    • A. 

      True

    • B. 

      False

  • 17. 
    Validity generalisation leads to differential validity OR test bias across groups. This is:
    • A. 

      True

    • B. 

      False

  • 18. 
    The SEE is used to determine the accuracy of that prediction. This is:
    • A. 

      True

    • B. 

      False