Chapter 4

Question 0

Alternate-form reliability (delayed administration)

Answer 0

two forms of the same test are administered on different occasions.

Question 1

Alternate-form reliability (simultaneous administration)

Answer 1

when two forms of the test are administered on the same occasion.

Question 2

Split half reliability

Answer 2

administering a test and dividing it into two equivalent halves that are scored independently.

Question 3

Coefficient Alpha and Kuder Richardson reliability

Answer 3

measures internal consistency.

Is sensitive to error. introduced by content sampling, and reflect the heterogeneity of test content.

Question 4

Composite Score

Answer 4

Reliability of composite scores is generally greater than the measures that contribute to the composite.

Question 5

Standard Error Measurement (SEM)

Answer 5

The standard deviation of the distribution of scores that would be obtained by one person would be tested on an infinite number of parallel forms of a test of items randomly sampled from the same content domain.

Question 6

Confidence Interval

Answer 6

reflects a range of scores that will contain the individuals true score with prescribed probability.

Question 7

Reliability

referred to as estimated

Answer 7

consitency or stability of assessment results.

Considered to be characteristics of results not the test itself.

Question 8

Measurement error

Answer 8

a degree of error, and some degree of error is present in all measurement.

Question 9

Test-retest Reliability

Answer 9

primarily sensitive to measurement due to time sampling and is an index of the stability of scores over time.

Question 10

Inter-rater Reliability

Answer 10

If the scoring of an assessment relies on subjective judgement, it is important to evaluate the degree of agreement when different individuals score the test.

Question 11

Kuder-Richardson Formula 20 (KR-20)

Answer 11

used when test items are scored dichotomously, that is simply right or wrong, scored 0 or 1.

*This is the most commonly used formula.

Question 12

Classical test theory

also referred to as “TRUE SCORE THEORY”

Answer 12

every mental test is composed of two components: True score (score if there were no errors) and error score.

x=T + E

Question 13

X= T+ E

Classical Test Theory

Answer 13

X (observed score) i (individual) = T (individuals true score true skills) with no error + E (measurement error)

Question 14

Is their a relationship between True Score and Error score?

A: Strong B: Weak C: No D: Moderate

Answer 14

No there is zero correlation or systematic relationship between true score and error score. True Score is without error

Question 15

True Score

Answer 15

True score has no measurement error. It reflects the individuals true skills, knowledge, abilities, attitudes whatever test is measuring.

Question 16

Random Measurement Error

Answer 16

The result of chance factors that increase or decrease an individuals score. (Classical Test Theory)

Question 17

Systematic Measurement Error

Answer 17

Systematic or consistent differences in test performance. Same amount of error every time. Associated with “TEST BIAS”.

Question 18

Content Sampling

Answer 18

Largest source of error in test scores.

Question 19

Time Sampling Error

Answer 19

Hunger, sleepy, sick, etc. Random fluctuations of performance

Question 20

Inter-rater (interscorer) differences

Answer 20

Subjective judgement testing, clerical error, administrator error.