Split-Half Reliability

  • Citations
  • Add to My List
  • Text Size

  • Entry
  • Entries A-Z

  • A statistical method for testing the reliability of a psychometric measure. In split-half reliability, all test items of a measure are randomly divided into two sets, which are then administered to participants. For each participant, the scores for each half are then calculated and correlated with each other. The split-half reliability estimate is based on the correlation between these two total scores. Some limitations of split-half reliability are related to the division of items, that is, how to determine the best criterion by which the items should be split. For example, a simple top half, bottom-half split is often not sufficient, due to practice effects, tiredness, and monotony. Other methods of testing reliability include the test-retest method and Cronbach's alpha.

    See also

    Looks like you do not have access to this content.


    Don’t know how to login?

    Click here for free trial login.

    • [0-9]
    • A
    • B
    • C
    • D
    • E
    • F
    • G
    • H
    • I
    • J
    • K
    • L
    • M
    • N
    • O
    • P
    • Q
    • R
    • S
    • T
    • U
    • V
    • W
    • X
    • Y
    • Z

    Back to Top

    Copy and paste the following HTML into your website