Revised short screening version of the attachment questionnaire for couples from the German general population.

In order to ensure the concentration, compliance and motivation of the participants, a short version of 12 items was extracted from the "Experiences in Close Relationships Questionnaire (ECR)". Even though this short English version shows equally good validity and reliability as the long version, there have been no representative norm values and psychometric characteristics available for the short version in German. Therefore, the German 12-item ECR was implemented in a representative sample of 1,127 males and 1,237 females (mean age M = 49.93; SD = 12.31) from the general public (N = 2,364). The reliability values of the German 12-item ECR in the representative sample are not as good as the long version with the 36- items version (Alpha = .54-72), and the 12-item ECR factorial structure failed to show the factorial validity. Since the EFA revealed that only half the items loaded on the expected factors, an even shorter form with six items was construed and tested psychometrically. Even though the item numbers were reduced, the reliability values of the German 6-item ECR improved and were as good as the long version with 36 items (Alpha = .73-90). Furthermore, factorial validity could be shown by CFA (CFI = .981, SRMR = .038, RMSEA = .079, TLI = .964) with scalar invariance across gender and age. In sum, the 6-item ECR is a reliable and factorial scalar attachment questionnaire. Due to its shortness, it is applicable to different research fields. However, reference data from a clinical sample are still missing.