Home › MCQs › Education › Question #6751
Back to Questions
Education QUESTION #6751
Question 1
A test has a Kuder-Richardson reliability of 0.85. A parallel form of the same test yields a test-retest-with-equivalent-forms correlation of 0.68. Which of the following best explains this discrepancy?
  • KR-20 measures stability over time, so it should always be lower than equivalent-forms reliability
  • KR-20 measures internal consistency within a single administration, while the equivalent-forms method captures both stability over time and equivalence across forms; thus the latter is typically lower✔️
  • The second test was harder, which automatically reduces reliability estimates
  • KR-20 is only applicable to essay tests, making the comparison invalid
Correct Answer Explanation
KR-20 is an internal consistency measure computed from a single test administration — it cannot capture the variance introduced by time passage or form differences. The equivalent-forms-with-retest method measures both stability and equivalence, capturing additional sources of variance that lower the coefficient. This explains why KR-20 (single administration) tends to exceed test-retest-with-equivalent-forms reliability.