5. A test is developed to measure how quickly people can type. A suitable method of assessing reliability would be:

a)Cronbach's alpha

b)Split-half

c)Item response theory

d)Test-retest

4.Classical test theory assumes that measurement errors

a)Are not correlated with true scores

b)Usually penalise the respondent

c)Are correlated from one test to another

d)Can be determined for the individual

3.In the context of self-report measures or tests, reliability could NOT be thought of as:

a)How precise the measurement is

b)Consistency of responses on a test

c)Confidence that a test is measuring what it really should be

d)Repeatability of measurement

2.To assess the reliability of a 120-question exam, an educator plans to use the Split-half method. How should this be done?

a)The questions in the first half of the exam should be compared to the second half

b)Cronbach's Alpha should be used instead of Split-half

c)The educator should use an Alternate forms approach

d)Odd and even numbered questions should be compared to each other

1.Based on the suggestion of Nunnally (1978) a suitable range for reliability could be considered.

a)-1 to +1

b).60 to at least .80

c).70 to at least .90

d)0 to 1