Verbatim · Inter-coder reliability
Krippendorff's unitizing alpha (uα)
DIFFERENT problem from fixed-unit alpha: here the coders independently SEGMENT a continuum, so the unit BOUNDARIES vary, not just the codes. Computed from the existing anchored coding spans (char_start/char_end) — each coder's coded spans are their units, the uncoded matter is gaps. The headline is the no-metric coefficient (unit IDENTIFICATION agreement); it answers 'did the coders mark the SAME stretches of text as relevant', not 'did they give them the same label'.
uA = UD = 1 - UDo/UDe, the UD-agreement over identified units on a continuum (no-metric metric delta_2=0: agreement on WHERE the units are, independent of how they are labelled)
- Measurement level
- no_metric
- Coders
- Two or more observer segmentations of a common continuum.
- Seminal reference
- Krippendorff (1995a), Sociological Methodology 25:47-76; corrected in Krippendorff, Mathet, Bouvry & Widlocher (2016), Quality & Quantity 50:2347-2364; canonical text: Krippendorff (2013/2015), 'Agreement when Unitizing and Coding Finite Continua', replacement of Section 12.4 in Content Analysis 3rd ed., Sage
Benchmark bands
Benchmarks are reporting guidance, not strict cut-offs — the engine presents them with their source.
| Agreement | Reading | Source |
|---|---|---|
| 0.80 – 1.0 | acceptable | Krippendorff (2013/2015), Section 12.4.2 p.314 — the same alpha-family standard as alpha>=.80 (Krippendorff 2004, p.241): rely on it. |
| 0.67 – 0.80 | tentative | Krippendorff (2004), p.241 — draw tentative conclusions in the .667-.80 range; applied to the alpha-family uA (Krippendorff 2013/2015 Section 12.4.2). |
| -1.0 – 0.67 | insufficient | Krippendorff (2004), p.241; Krippendorff (2013/2015), Section 12.4.2 — below the .667 alpha-family floor. |
Verbatim computes Krippendorff's unitizing alpha (uα) with a deterministic, golden-verified math kernel — never a language model — with a bootstrap confidence interval and full provenance (coder + unit + kernel + fixture + run-id). The number is measured; the interpretation stays yours.
Measure this in the workshop →