Week 4 Lecture

Question 1

________ (consistency, repeatability)
_______ (measures what it is supposed to measure)
__________ (quality of normative data)

Answer 1

Reliability; Validity; Standardisation

Question 2

R_______ of a measure – extent of being free from measurement error

Answer 2

Reliability

Question 3

______ _____ ______ theory
● Observed score = True score + Measurement error
● Errors assumed to be random
● Standard deviation of the distribution of errors shows magnitude of measurement error

Answer 3

Classical test score

Question 4

Test-retest reliability correlation between scores of repeated administration (r = coefficient of _______)

Answer 4

stability

Question 5

What is an indicator of reliability for nominal data?

Answer 5

KR20 (Kuder-Richardson 20)

Question 6

What statistic is used for inter-rater reliability?

Answer 6

kappa

Question 7

What is satisfactory reliability for research purposes?

Answer 7

.7 - .8

Question 8

What is satisfactory reliability for tests affecting someones future?

Answer 8

.95 ( r > = .85 for comparing individuals with others)

Question 9

Derive uni-dimensional tests through f_____ a_______. This will reduce content
heterogeneity. This is a way to increase reliability

Answer 9

factor analysis

Question 10

_______ validity
– items/questions need to relate to behaviour trait
– inter-rater agreement of items

Answer 10

content validity

Question 11

________ validity

– Meeting some level, i.e. predictive ability

Answer 11

Criterion validity

Question 12

________ validity
– Theory driven; usually a quality or trait that is difficult to define (may be several
characteristic).

Answer 12

Construct validity

Question 13

__________ validity: high correlations with measures of similar constructs

Answer 13

Convergent validity

Question 14

___________ validity: low correlations with

measures of unrelated constructs

Answer 14

Discriminant validity

Question 15

__________ validity (simultaneous), e.g., correlating with an established and/or more elaborate test, job samples, diagnostic tests

Answer 15

Concurrent validity

Question 16

The validity coefficient is examined by the correlation between the test and a specific c________

Answer 16

criterion

Question 17

R2 (or v______ c________ squared) is the coefficient
of determination; variation in the criterion explained
by the test scores

Answer 17

validity coefficient

Question 18

______ ______ of Estimate

- Margin of error to be expected in the individual’s predicted criterion score

Answer 18

Standard Error

Question 19

Standard Error of Estimate
With perfect validity or (R12)2 = _,
standard error of estimate = _.

Answer 19

1;0

Question 20

______________ tests usually have considerably higher reliabilities than ___________ tests

Answer 20

Intelligence tests; Personality

Question 21

Does personality or intelligence tests have more reliability?

Answer 21

Intelligence