Worldview LensHelp calibrate the models
Worldview Lens is checked against certified human raters for ego development, where readings now land within one step of an expert rating 91% of the time. The other models have no expert check yet. If you hold texts scored by trained raters, you can change that.
What we're looking for
Texts written by real people, each with the score a trained rater gave it. About 30 scored texts per model is enough to start, ideally spread across several stages. More is better, and item-level scores help as well as overall ones.
| Model | Scored with | Status |
|---|---|---|
| Kegan | Subject–Object Interview transcripts with their scores | Wanted |
| Kohlberg | Moral Judgment Interview responses with stage scores | Wanted |
| Fowler | Faith Development Interview transcripts with stage scores | Wanted |
| Commons (MHC) | Texts or task responses scored for order of hierarchical complexity | Wanted |
| Torbert | Leadership development profiles (sentence completions) with action-logic ratings | More welcome; first fit from SCTi data |
| Cook-Greuter, Loevinger | SCTi-MAP or WUSCT protocols with item ratings and TPR | Calibrated; more data tightens it |
Spiral Dynamics and Wilber's altitudes have no validated way of scoring writing, so they can't be calibrated this way; How scoring works says so. We don't need scoring manuals, only texts and the scores they were given.
How your data is treated
- Only with consent. Share texts only where the people who wrote them agreed to research use, under your own ethics rules.
- De-identified first. Remove names, places and anything that identifies the writer before sending.
- Private. Kept off the public site and out of our code repository, used only to calibrate and test the scorer, never published, sold, or used to train other systems.
- Results only in aggregate. What appears on How scoring works is summary agreement, like the ego-development results there.
- Yours to withdraw. Ask and we delete it. We're glad to sign a simple data agreement first.
What you get
- A free year of Worldview Lens Coach for a usable set of scored texts.
- A private comparison of how the scorer reads your texts against your raters, before anything is published.
- Credit on How scoring works, by name or anonymously, as you prefer.
How to start
Email support@worldviewlens.com with the subject Calibration data. Say which model and instrument, roughly how many scored texts you have, and how they were consented. Please don't attach data in the first email; we'll agree a secure way to share it.