Methodology
How we score your writing voice
The Voice Profile is a stylistic mirror, not a psychological assessment. It describes patterns in your writing — sentence length, word choices, how often you hedge vs. assert. It says nothing about your intelligence, personality, or ability.
What we analyse
When you paste emails, we strip everything that isn’t your own writing before scoring: quoted reply lines (lines starting with >), “On [date] … wrote:” headers, forwarded-message separators, and common signature blocks. Only the sentences you actually authored are counted.
The four scores
Each score runs 0–100, calibrated so 50 ≈ typical professional email. These baselines are estimated from general business-writing norms, not empirically validated against a representative corpus — treat them as a relative compass, not an absolute ruler.
How often your writing uses assertive, authority-signalling language versus tentative, hedging language.
We count how often you use words like always, must, confirm, lead, require — authority signals — versus words like maybe, perhaps, believe, feel, roughly — tentativeness signals. Net rate, normalised to 0–100.
Inspired by the LIWC 2015 Clout and Tentativeness categories (Pennebaker et al., 2015).
How often your writing uses positive-affect, appreciative, and inclusive language.
We count words like thanks, appreciate, wonderful, we, us, our, team, celebrate — positive-affect and social-inclusion signals. Rate normalised to 0–100.
Inspired by the LIWC 2015 Positive Emotion (Posemo) and Social categories (Tausczik & Pennebaker, 2010).
The ratio of certainty markers to uncertainty markers in your writing.
We count booster words (definitely, certainly, clearly, always, indeed…) and hedge words (might, may, perhaps, seem, approximately, arguably…) and compute their ratio. Equal boosters and hedges = 50; more boosters = higher conviction.
Inspired by Hyland’s (2005) hedge and booster framework for professional writing (Hyland, K., Metadiscourse, Continuum).
How readable your sentences are — shorter sentences and simpler words score higher.
We apply the Flesch Reading Ease formula: 206.835 − 1.015 × (words ÷ sentences) − 84.6 × (syllables ÷ words), clamped to [0, 100]. Higher clarity means shorter sentences and more common words — not simpler thinking.
Flesch, R. (1948). A new readability yardstick. Journal of Applied Psychology, 32(3), 221–233.
How your archetype is assigned
Archetypes are assigned by a deterministic rule set that matches your four scores — plus your average words per email — against a ranked list of patterns. The first matching rule wins. For example, a high Clout + high Warmth writer becomes “The Velvet Hammer”; a very short average length + high Clarity becomes “The Sprinter.” The full rubric is versioned (RUBRIC_VERSION = “v1”) so results are stable and reproducible.
The radar on your card plots your four dimensions — Clout, Warmth, Conviction, Clarity — against the average writer, so the shape at a glance is your voice.
What we don't claim
- Our word lists are proxies inspired by the cited frameworks — not licensed replications or certified implementations.
- The calibration constants are estimated from general norms, not empirically validated against a large representative corpus of professional email.
- Scores describe patterns in the text you pasted, not a fixed trait. Write more formally one week, and the scores will shift.
- This is not a clinical, psychological, or personality assessment. Do not use it to make hiring, promotion, or health decisions.