Adaptation (test theory)
When adapting a psychological test to a different culture, in addition to being translated into another language, the test must also be re-examined and re- standardized with regard to its test quality criteria . This quality assurance in psychological diagnostics is necessary because the meanings of formulations and symbols as well as the distribution of statistical parameters can differ from one another in different cultures. The International Test Commission (ITC) published a guideline for this process, which in 2000 still bore the title "Test Translation and Adaptation Guidelines" (TTAG) and later in 2001 was only called "Test Adaptation Guidelines" (TAG) has been. The TAGs have been used for important studies such as the Third International Mathematics and Science Study (TIMSS) and the PISA studies .
The aim of an adaptation is that a test that has been translated into another language or adapted to another culture covers an aspect equally in two or more cultures, i.e. are equivalent in the sense of Geisinger (2003) . According to Geisinger, four types of equivalence can be distinguished:
- Linguistic equivalence: This is not a literal translation, but a translation of the meaning. The saying: “Compare apples with pears” should accordingly be translated into: “to compare apples and oranges”. Images and characters also have to be translated. Since unfamiliar material means additional thought for the test persons, images have to be adapted, such as people's culturally customary clothing.
- Functional equivalence: This refers to the fact that aspects in tests must have the same purpose and have the same meaning. While in Germany praise can be expressed by forming a ring with your thumb and forefinger, in Italy this symbol would be taken as an insult.
- Conceptual equivalence: This equivalence exists when the same construct is recorded in both cultures. This is the case when the same nomological network exists in the cultures and the same relationship can be shown between the measured construct and other constructs in both cultures (divergent and discriminative validity).
- Metric equivalence: The metric equivalence is given if the tests agree in their statistical parameters. For example, the items (tasks or questions) should be identical in their item statistics (comparable distribution of answers and comparable task difficulty) and the reliability should not show any difference. The conceptual equivalence is also a prerequisite for this.
Test Adaptation Guidelines
The Test Adaptation Guidelines consist of four sections:
- Section 1: This deals with guidelines on construct equivalence, i.e. the question of whether the construct that is to be recorded by the test actually exists in both language cultures.
- Section 2: Guidelines for translating questions, collecting data and checking test quality criteria are given here. For example, at least two translators should be involved who have knowledge of both language cultures.
- Section 3: This section deals with what needs to be considered when carrying out the test with groups that differ linguistically and culturally. The instructions relate to the choice of test users, the choice of test items and the time restrictions.
- Section 4: At the end it is pointed out that good test documentation is necessary.
Individual evidence
- ^ Franz Petermann, Michael Eid : Handbuch der Psychologische Diagnostik . Hogrefe Verlag, 2006, ISBN 978-3-8409-1911-4 , pp. 411 ( limited preview in Google Book search).
- ↑ a b c d e f g h Helfried Moosbrugger, Augustin Kelava: Test theory and questionnaire construction . Springer-Verlag, 2011, ISBN 978-3-642-20072-4 , pp. 210–211 ( limited preview in Google Book search).
- ↑ a b c d e f g h i j k l Lutz F. Hornke, Manfred Amelang, Martin Kersting, Niels Birbaumer, Dieter Frey: Subject area B: Methodology and methods / psychological diagnostics / basic questions and fields of application of psychological diagnostics . Hogrefe Verlag, 2011, ISBN 978-3-8409-1523-9 , pp. 211–212 ( limited preview in Google Book search).