Reliability (psychometrics)

The reliability (German: reliability) is a measure of the formal accuracy or reliability of scientific measurements. She is the one fraction of the variance that can not be explained by measurement errors caused by actual differences in the measured characteristic and. Hochreliable results must be substantially free of random errors, ie when repeating the measurement under the same conditions, the same measurement result would be achieved ( replicability of results under the same conditions ).

The reliability is in addition to the validity and objectivity of the three most important quality criteria for empirical studies dar. High reliability is basically a requirement for high validity, with a high reliability can be at the expense of validity ( Reliabilitäts - Validitäts dilemma).

Reliability includes three aspects:

• Stability (equal or similarity of the measurement results when used at different times )
• Consistency ( the extent by which all items that are summarized in a test for a characteristic feature measure the same thing )
• Equivalence ( equivalence of measurements)

In the psychological assessment it is considered one of the main quality criteria of psychological tests. It specifies exactly how a personality or behavioral trait is measured.

Types

The reliability can be estimated by different methods. Depending on the method is spoken by other Reliabilitäts types.

Possibilities of Reliabilitätsverbesserung

• The reliability of the tests can be improved by extending the tests using similar items, because with the length of the test increases the measurement accuracy.
• The objectivity is a necessary condition for the reliability. Accordingly, an objective improvement can increase the reliability of the measuring instrument.
• In formulating the items item homogeneity should be sought. Items are homogeneous if they contain each other. This means that subjects who answered in the affirmative the most extreme item, also affirm the weaker formulated Item or deny a negatively poled item.
• Little separated sharp items should be excluded. An item that separates well between people with low and high characteristic value, contributes to the accuracy of the test.