Week 2 - Mandy's Notes

Question 1

Define validity

Answer 1

Textbook: the extent to which evidence supports the meaning and use of a psychological test (or other assessment device)

Mandy’s words: The degree to which a measure of a concept truly reflects that concept
» Does the test do what it is designed to do?

Question 2

Explain the relationship between reliability and validity

Answer 2

A test can be reliable without being valid. But, to be valid, a test must be reliable

Question 3

Criterion – predictive

Answer 3

How well test predicts future performance
- Check correlation between scores and criterion in the future. Results in more error than concurrent as events can intervene between the time of test and the criterion.

Question 4

Construct

Answer 4

Test reflects underlying construct
- Correlate test scores with other established, valid tests of similar constructs (or distinct constructs)
• Demonstrate that test distinguishes between people
• Check task requirements test the components of the underlying theory
- Can include convergent and divergent validity

Question 5

Convergent (within construct validity)

Answer 5

tests whether concepts or measurements that are supposed to be related are actually related.

Question 6

Discriminant validity (or divergent validity) (within construct validity)

Answer 6

tests whether concepts or measurements that are not supposed to be related are actually unrelated.

Question 7

How dis Diener et al (1985) create a criterion validity coefficient?

Answer 7

Diener and colleagues (1985) obtained a criterion validity coefficient for the SWLS in terms of life satisfaction rating made by experimenters who interviewed each subject about their life.

Question 8

Validity in Clinical Practice (two types)

Answer 8

Incremental validity and conceptual

Question 9

Incremental validity

Answer 9

Does the test produce accurate results above and beyond what could be obtained with greater ease / less expense. i.e could evaluate the person’s years of education to predict performance on a job instead of test

Question 10

Conceptual Validity

Answer 10

Hypothesise about the individual and gather data from multiple sources
See if it all matche sup–is it logical? Is the information contradictory?

Question 11

Standardisation sample of WISC-V (6pt)

Answer 11

1- N = 2200 children from the U.S aged 6:0 to 16:11
2- » Divided into 11 age groups.
3- Equal number Ma and F each grp
4- Ethnic proportions representative of the population in the sample
5- Five parent education levels included
6- Representation from states across US.

Question 12

WISC-V Reliability (what did they obtain with split half method?)

Answer 12

1) Internal consistency obtained using split-half method with normative sample
» Average reliability coefficient subtests range from .81 to .94
» Reliability coefficients for primary index scores range from .88 to .93

Question 13

WISC-V Reliability Test-retest details (5pt)

Answer 13

Test-Retest Stability
1- Sample of 218 children
2- Participants administered WISC-V twice,
3. with test-retest intervals ranging from 9-82 days, and a mean interval of 26 days.
4. - Test-retest reliability estimated using Pearson’s product-moment correlation
5- Scores demonstrate adequate stability across time for all ages

Question 14

WISC-V Reliability

» Interscorer /Inter-raterReliability

Answer 14

All protocols double-scored by two
independent scorers
» Evidence of agreement obtained using normative sample
» Interscorer agreement very high (.98 - .99)
36

Question 15

WISC-V Validity - How did they establish Content validity?

Answer 15

Content validity established via comprehensive literature and expert and advisory panel reviews

Question 16

WISC-V Construct validity established via: (6pt)

Answer 16

1) examination of response processes during scale development
2) » Interrelationships among subtests
3) » Factor analysis – establishing psychological meaningfulness
4) » Relationships to other measures
5) » Demonstrate that test distinguishes between people
6) » Assessment of children from special clinical and educational populations

Question 17

WISC-V Validity

Criterion validity established via (1pt)

Answer 17

1) Relationships to other measures with established validity

» E.g.,WISC-IV

Question 18

Criterion Concurrent

Answer 18

Check correlation between scores and a criterion in the present.

Question 19

Standard error

Answer 19

• larger the sample, the less error, the more reliable and valid. SEM = √ SD ( 1 - r )