Interbedömarreliabilitet – Ett Tillförlitligt Mått På Standardiserade Intervjuer?

Abstract

Aims Inter-rater reliability studies are unusual as regards the standardised interview instruments used in the Swedish treatment care system for substance abusers. This article presents the results of a study intended to establish the inter-rater reliability of DOK interviews, through identification of question areas and items with high or low correspondence, and through analysis of how ratings differ between different interviewers. Methods & Data In total 41 practitioners participated in this study, with various experiences and from six different treatment units of the Swedish treatment care system for substance abusers. Participants gave their ratings and completed the DOK questionnaire based on a video-taped interview. Calculation and analysis of the correspondences were carried out for each individual question area and also each individual item by using appropriate statistical methods. Differences between the various assessors are presented. Results The results of the study show that the percentage agreement for the question areas and the items is generally high, with the exception of a few isolated questions. Conclusions Among the many possible different explanations for a lack of correspondence is that the questions could be ambiguously constructed and defined or the interview guidelines do not offer satisfactory guidance, or some questions could be too detailed. In addition, a few of the raters may lack sufficient training or experience when it comes to conducting interviews based on DOK. The combined results of the study show that there is a high level of inter-rater reliability, which indicates that in practice it is possible to utilise structured interviews such as DOK.