1-5 Flashcards
What are the three types of reliability?
test-retest
internal consistency
classical test theory (CTT)
test-retest reliability
estimate the reliability/precision of test scores when the same people take the same test form on two separate occasions.
estimate the reliability/precision of test scores when the same people take the same test form on two separate occasions.
test-retest reliability
internal consistency
estimate the reliability/precision of test scores by looking at the relationships between different parts of the same test given on a single occasion
estimate the reliability/precision of test scores by looking at the relationships between different parts of the same test given on a single occasion
internal consistency
How do we establish internal consistency for a test?
> >
KR-20
formula for calculating internal consistency, for right or wrong answers (multiple choice, t/f)
formula for calculating internal consistency, for right or wrong answers (multiple choice, t/f)
KR-20
Coefficient alpha (Cronbach’s alpha)
calculates internal consistency for questions that have more than two possible responses such as rating scales
calculates internal consistency for questions that have more than two possible responses such as rating scales
Coefficient alpha (Cronbach’s alpha)
When is it appropriate to use coefficient alpha to determine an estimate of internal consistency reliability?
calculate and report an estimate of internal consistency for each homogeneous subtest or factor
NOT heterogenous
Classical test theory (CTT)
a person’s test score (called the observed score) is made up of two independent parts: true score (T) and random error
a person’s test score (called the observed score) is made up of two independent parts: true score (T) and random error
classical test theory (CTT)
true score (T)
The score that would be obtained if an individual took a test an infinite number of times and then the average score across all the testings were computed
The score that would be obtained if an individual took a test an infinite number of times and then the average score across all the testings were computed
true score (T)