Читать книгу The Handy Psychology Answer Book - Lisa J. Cohen - Страница 245

What does it mean to say a test is reliable?

Оглавление

The reliability of a test refers to its ability to measure a given trait consistently. If the outcome of a measure varies each time it is applied, the measure is not reliable. There are several forms of reliability, depending on the format and purpose of the test. Inter-item consistency means that the individual items of a test are inter-correlated, or they are well related to each other. This form of reliability is used with questionnaires in which multiple items are used to rate one trait. Test-retest reliability measures how well an initial administration of a test correlates with a repeated administration. This is only useful if the trait measured is unlikely to change much over time. Inter-rater reliability is used with semi-structured questionnaires and other instruments in which the rater must use complex subjective judgments in the scoring. An instrument has inter-rater reliability when two or more raters rate the same material the same way.

The Handy Psychology Answer Book

Подняться наверх