Week 3 Many's Notes

Question 1

What are different ways of expressing test scores?

Answer 1

The raw score is expressed in terms of the DISTRIBUTION of a sample of individual’s scores when NORM-REFERENCING.

Question 2

What’s the difference between norm referencing and criterion referencing in interpreting test scores?

Answer 2

Criterion-referenced tests: the task itself is a yardstick to which performance is reffered, so the raw score has meaning in itself ( e.g. reading test, driving test)

Norm referenced: Raw score is compared to the average or norm of a representative group of people similar to the person being tested. More common in psych testing.

Question 3

What are different methods for transforming scores?

Answer 3

We start with a set of test items i.e. test items, and we have their item scores. Item scores are calculated to give a total raw score. As the raw score is often meaningless, we norm-reference them (compare them to norm).

Question 4

What is meant by standard score?

Answer 4

Percentage of cases below a z of +1 in normal distribution. Not to be confused with a STANDARDISED SCORE for subtests: M set at 10, SD = 3

Question 5

How is a standard score interpreted?

Answer 5

We think… it is the percentage of cases below z of +1 in a normal distibution

Question 6

What is meant by a percentile?

Answer 6

The percentile scale expresses raw scores in terms of the percentage of cases that lie below it. e.g. IQ in the 98th percentile means 98% lie below it. Not % correct

Question 7

Deviation IQ

Answer 7

Because Z scores are untidy (+/-) Wechsler (1955) set the mean at 100 and set the SD at 15 and adjusted the scores.

Question 8

Examples of transformed z scores

Answer 8

T score:Mean is set at 50, SD at 10 (e.g.,
MMPI)
Sten score: Mean is set at 5.5., SD at 2 (16PF)

STANDARDISED SCORE for subtests: M set at 10, SD = 3

Question 9

When transforming raw scores to z scores what do we need?

Answer 9

We need the mean and SD of the comparison group

Question 10

If comparison is not made to appropriate group:

Answer 10

The transformation would fail to convey meaning

Question 11

When is it useful for norms to be age based?

Answer 11

If the characteristic varies across the lifespan. Other tests might use norms to compare a community sample vs a clinical sample.

Question 12

What is an adequate sample size for a norm? In fact what consititutes excellent, and reasonable and the others?

Answer 12

2000+ is excellent
1000-2000 is good
500-1000 is reasonable
200-500 is adequate
Under 200 is inadequate

Question 13

What was the sample size for the testing of the WAIS - IV?

Answer 13

N = 2200

Question 14

What was the sample size for the MMPI?

Answer 14

N=2600 (convenience) 7 US states

Question 15

What was the sample size for the 16PF?

Answer 15

2500, stratified by age, gender, education level, race

Question 16

According to O’Shum et al., 2017, pp. 63-64), what are the considerations to be made on norms? (6)

Answer 16

1.Check SOURCE
2.Are norms RELEVANT to this situation?
3.If concerned about country of origin, what is known
about susceptibility of test to CULURAL differences?
4.How is the test RESULT being used?
5.Is it possible to check in another way against ANOTHER TEST with appropriate norms? (beware small samples)
6.EXPLAIN in your RESPORTS if there are norm issues and that results should be interpreted with caution

Question 17

What is a reading age?

Answer 17

A way of providing age and grade equivalents.

Age 10 Reading Age = MEDIAN reading score for children aged 10