Split-half reliability is estimated by which method?

Split-half reliability is estimated by which method?

• A. Correlating two halves of the test and applying a correction.
• B. Re-administering the test to the same participants on a different day.
• C. Comparing pilot data to full data.
• D. Comparing means of two independent samples.

Answer: (A)

Split-half reliability measures internal consistency by checking whether items on the same test hang together as a single construct. To estimate it, you divide the test into two halves (for example, odd vs even items), compute each participant’s score on both halves, and then correlate the two sets of half-scores. Because halves are shorter, you apply a correction (the Spearman-Brown formula) to adjust that correlation so it reflects the reliability of the full test. This approach targets whether the test items consistently measure the same attribute. Re-administering the test on a different day assesses stability over time (test-retest), not internal consistency; comparing pilot data to full data and comparing means of two independent samples address different questions altogether.

Split-half reliability measures internal consistency by checking whether items on the same test hang together as a single construct. To estimate it, you divide the test into two halves (for example, odd vs even items), compute each participant’s score on both halves, and then correlate the two sets of half-scores. Because halves are shorter, you apply a correction (the Spearman-Brown formula) to adjust that correlation so it reflects the reliability of the full test. This approach targets whether the test items consistently measure the same attribute. Re-administering the test on a different day assesses stability over time (test-retest), not internal consistency; comparing pilot data to full data and comparing means of two independent samples address different questions altogether.