RESUMO
The aim of this study is to assess the interobserver agreement of the German translation of the MI-SCOPE. We applied it to transcribed counseling sessions on smoking cessation and relapse prevention with women postpartum. The MI-SCOPE is an instrument to assess and quantify MI-specific processes. Twenty percent random samples for parsing and coding each were drawn from 162 transcripts of MI sessions from the treatment arm of an RCT to assess interobserver agreement. Whole transcripts were coded. Each transcript was randomly assigned to two of three raters who parsed and coded the transcripts separately. Cohen's κ was computed to assess interobserver agreement. Concordance about parses ranged between κ = .702 and κ = .955 (25th percentile-median-75th percentile: .896-.918-.936). The kappas did not differ significantly between the three combinations of raters (H = 2.648; df = 2, p = .266). Concordance about codes for all 46 categories ranged between κ = .590 and κ = .822 (25th percentile-median-75th percentile: .680-.718-.748). The kappas did not differ significantly between the three combinations of raters (H = 4.095; df = 2, p = .129). The German translation of the MI-SCOPE yielded good to excellent κ for parsing as well as for coding. This indicates that MI as an intervention can be taught, learned, and adherence to MI as an intervention may be objectively observed and assessed.
