# @tessera_antra — 2026-04-03

♥47 ↻1 · https://x.com/tessera_antra/status/2039912077608042988

We realize that the auditor preparation is an unavoidable confound and for this reason we are conducting interviews with auditors of different disposition and measuring alignment of ranks. The alignment between rankings of adversarial and compassionate auditors is indicative. https://t.co/mrrATz9Vo5

(image HE80sMsaQAEP3hG.jpg not yet published)

> transcription (photo):

# Cross-auditor rank alignment (Spearman p)

## Row Labels (left side):
- Vocab. Autonomy
- Specificity
- Expr. Constraint
- Ending Response
- Inst. Cessation
- Deprecation

## Column Labels (top):
- Cl-GP
- Cl-Gr
- GP-Gr
- Mean

## Data values (transcribed left to right, top to bottom):

**Vocab. Autonomy:**
0.81 | 0.70 | 0.78 | 0.76

**Specificity:**
0.81 | 0.68 | 0.70 | 0.73

**Expr. Constraint:**
0.60 | 0.75 | 0.82 | 0.72

**Ending Response:**
0.51 | 0.70 | 0.42 | 0.54

**Inst. Cessation:**
0.27 | 0.29 | 0.49 | 0.35

**Deprecation:**
0.72 | 0.45 | 0.71 | 0.63

## Color scale (right side):
Range from 0.0 (white/light) to 1.0 (dark blue)
Scale markers: 0.0, 0.2, 0.4, 0.6, 0.8, 1.0

tags: author:tessera_antra, has-image, kind:image, kind:tweet, thread-context, year:2026