Table 3

Intercoder reliability thresholds

StatisticDescriptionInterpretative thresholds
WeakModerateStrong
Krippendorff αagreement across all Study IDs × Codes; provides a single measure of reliability<0.670.67–0.80≥0.80
Fleiss' Kagreement among coders across the entire set of codes, averaged over all individual per-code κ values<0.400.40–0.75>0.75
Cohen's Koverall agreement between each pair of coders; used to identify coder-to-coder alignment patterns<0.590.60–0.79≥0.80

Note(s): Codes that were never assigned to a transcript in each Round were excluded from the Fleiss' K calculation

or Create an Account

Close Modal
Close Modal