# Measurability Notes — Wave 1 Term Set

> **Living Document** — refinable within the Refinement Window (Disclaimer V6 §31). Definitions stable under CC BY-NC-ND 4.0. This file is part of one observer's research artefact; readers are invited to build their own.
>
> **Lebendiges Dokument** — verfeinerbar im Refinement-Window (Disclaimer V6 §31). Definitionen stabil unter CC BY-NC-ND 4.0.

---

**Stand / Last updated:** 2026-05-27

These notes describe how the 104 Wave 1 terms can be observed and counted in everyday AI-interaction practice. They are descriptive measurement protocols, not normative claims; they invite empirical study but do not assert validation.

---

## Measurement Categories

### Category 1 — Subjective State Reports (Self-Rating)

Terms describing internal states (joy, calm, pride, etc.) lend themselves to a simple 1–7 Likert-style self-rating immediately after an AI interaction. Examples:

| Term | Self-Rating Item |
|---|---|
| **Pause Power (AUG-1005)** | "Before my next prompt I deliberately paused to improve my next question." (1 = not at all, 7 = strongly) |
| **Joy Jolt (AUG-1027)** | "I experienced a brief, unexpected moment of joy during this interaction." |
| **Heart High (AUG-1025)** | "I felt connected, joyful, or meaningful during this session." |
| **Serenity Spark (AUG-1012)** | "A small response triggered disproportionate inner calm." |
| **Pride Pulse (AUG-1022)** | "I felt a flash of pride from progress mirrored in this session." |
| **Day Highlight (AUG-1059)** | "This session contains a moment I'd mark as a day-highlight." |

A daily diary form with these items produces frequency counts per term per week — observable, comparable across users, replicable.

### Category 2 — Behavioural Markers (Observable)

Terms describing behavioural patterns can be observed via session logs:

| Term | Observable Marker |
|---|---|
| **Pause Power (AUG-1005)** | Inter-prompt latency > N seconds before next prompt (where N is user-calibrated) |
| **Termination Reluctance** (in compendium cluster) | Session length beyond stated task-completion |
| **Reformulation Dependency** | Ratio of user-edited-AI-output to user-original-output |
| **Vocabulary Colonization** | Frequency of AI-introduced terms re-used by user across sessions |

### Category 3 — Rhythm / Cadence Patterns

Terms describing temporal patterns:

| Term | Measurement |
|---|---|
| **The Vigilance Rhythm (AUG-0975)** | Verification-checkpoint frequency: how often the user pauses to verify within a session, vs. continuous unverified flow |
| **Calm Wave (AUG-1010)** | Self-reported calm trajectory across a session (1–7 rating at start, middle, end), or transcript-coding of calm-indicating language over the session arc |
| **Heart Reset (AUG-1064)** | Pause-and-re-center moments within a session, self-marked or coded from transcript markers |
| **Calm Reset (AUG-1043)** | Brief pause-and-re-center after intense AI interaction |

### Category 4 — Trust & Calibration Items

| Term | Measurement |
|---|---|
| **The Calibrated Trust (AUG-0422)** | Verification-effort allocated as a function of stakes (high-stakes vs. routine output) |
| **The Source Curiosity (AUG-0330)** | Frequency of "where does this come from?" follow-up queries per session |
| **The Companion Pattern (AUG-0981)** | Steady-companion integration vs. substitute-for-thinking pattern in daily logs |

### Category 5 — Humour / Repair Patterns

| Term | Measurement |
|---|---|
| **Loop Laugh (AUG-1004)** | Humour-loop instances per session (user laughter, light corrections, dramatised AI mistakes) |
| **Adjustment Aha (AUG-1037)** | "I realised the issue was my prompt, not the output" — self-marked or detected via prompt-edit-after-output |
| **Oops Moment (AUG-1030)** | Brief acknowledged AI misstep used as learning signal |

---

## Recommended Diary Protocol

A simple diary form for self-study:

1. After each substantial AI session, rate 5–10 most relevant terms on a 1–7 scale.
2. Keep entries for 14 days minimum to observe individual base-rate patterns.
3. Optionally pair with passive markers (session length, inter-prompt latency, prompt-edit-after-output counts from the transcript).

This is a research protocol invitation, not a clinical instrument.

---

## Limitations

- All measurements are subjective unless paired with behavioural log markers (e.g. session length, inter-prompt latency, transcript coding).
- No inter-rater reliability has been formally validated.
- Term identification across cultures and languages requires further research (Wave 2+).
- Falsifiability is per-term and protocol-specific.

---

## Related Files

- [Disclaimer V6 §1–§40](./DISCLAIMER.md) — §29 Empirical Claims Subject to Peer Review
- [Impressum / Legal Notice](./IMPRESSUM.md) — DDG §5 / MStV §18
- [Privacy Policy / Datenschutz](./PRIVACY.md) — DSGVO 9-Punkte
- [AI Authorship Disclosure](./AI_DISCLOSURE.md) — EU AI Act Article 50
- [augmanitai_wave1_104.jsonld](./augmanitai_wave1_104.jsonld) — full term definitions
- [augmanitai_wave1_sampler.csv](./augmanitai_wave1_sampler.csv) — tabular overview
- [README](./README.md) — overview

---

**License:** CC BY-NC-ND 4.0