Abstract
The objectives of the study were to develop a version of the Cambridge depersonalization scale (CDS) and to examine its reliability, validity and discrimination quality as an instrument for screening depersonalization disorder. Thirteen subjects met ICD-I O criteria for depersonalization disorder, 69 for anxiety disorder and 65 were healthy controls.
Cronbach's alpha and splithalf reliability was carried out to assess reliability. Spearrnan correlation analysis between CDS and DES depersonalization subscale and the Kruskal-Wallis test to compare scores ofthe CDS among the groups were carried out in order to examine criterion validity.
Analysis ofthe receiver operating characteristics (ROC) was perforrned to examine discrimination duality of the CDS as an instrument for screening depersonalization disorder. The version ofthe CDS showed high interna I consistency (Cronbachs alpha and split halfreliability were both 0.95) and convergent validity (Spearrnan correlation with DES depersonalization subscale was 0.68).
The scale was able to differentiate patients with ICD- 10 depersonalization disorder from the other groups. The area under the ROC curve was 0.975 and the best compromise between the true positive and false negative rate was at a cut-off point of 50, yielding a sensitivity of 1.00 and a specificity of 0.87.
The version ofthe CDS has good reliability, validity and discrimination quality, similar to the original instrument and can be used in both clinical research and practice. Results should be interpreted with caution given the relatively small Hample size of the depersonalization disorder group.
A larger sample in future study will be needed.