- Objectivity
- Reliability
Definition
Validity is defined as the degree of accuracy with which a test procedure actually tests the feature it is designed to measure. So… the degree to which the test measures exactly what it claims to measure. The validity is therefore an important scientific criterion of quality.
The question to be answered is: What is actually measured in the test/measurement procedure? does the test really measure exactly what is to be measured? Prerequisite: What is to be measured must be clarified in advance (definition of the ability or the construct).
In training-scientific performance diagnostics, the first two forms of validity are of primary interest, especially the criterion validity.
- Logical argumentation (content validity)
- Determination of the correlation with the criterion (criterion validity)
- Determination of the correlation with similarities (contract validity)
The validity of the content is given if validity is inferred from theoretical or logical considerations and not from empirical investigations. 1) Apparent validity: The validity of the content is based on the well-founded presumption of validity.
So… the test gives the impression, that it actually tests the ability, which fits to it. If the agreement with the characteristic is so obvious, one speaks also of logical, or trivial validity. For example (shots at soccer from the 16 m. line as determination of the shooting power or free throws from the free-throw line as determination of the throwing precision in basketball).
2. expert rating: If the validity of the content is not obvious, further investigations by competent persons are necessary. A test cannot be definitively determined as valid in terms of content. It must first be confirmed by expert opinions.
The expert rating is understood as a consensus of experts. Procedure of the expert rating in determining validity: The validity of the content is considered positive if more than 80% agreement is reached. Note!
(Problems of content-related validity)
- The selected experts are asked either in writing or orally whether the test checks the feature to be tested
- Experts receive a questionnaire and must tick the skills that will be determined during the test.
- A measuring instrument always tests several abilities and/or skills in parallel and technical differences in the test execution can falsify the result
- Functional Fluctuations (The same test may test different characteristics in different subjects. E.g. push-ups (strength endurance vs. maximum strength at different performance levels)
