Our Methodology

Saati labels every result as either a Validated Measure (WHO-5, PERMA-Profiler, used exactly as published) or a Saati Insight (our own informal reflection questions) — the two are never blended into a single unlabeled score.

Saati is not a medical or diagnostic service. It’s a wellbeing check-in built on a mix of established research instruments and our own evidence-informed questions — and we think you deserve to know exactly which is which.

Two kinds of measures, never blended

Every result in a Saati report is labelled as one of two types, and we never combine them into a single unlabeled number:

• Validated Measures— published, independently researched instruments we use exactly as their authors wrote them, with official scoring formulas we don’t alter.
• Saati Insights— questions we designed ourselves, informed by general research but not independently validated as a scored instrument. We’re upfront that these carry a different, lower evidentiary weight.

The validated instruments we use

WHO-5 Wellbeing Index — a five-item measure developed by the World Health Organization, used worldwide in both research and clinical settings as a short general wellbeing screener. We use the official wording and the official raw-sum scoring formula, unaltered.

PERMA-Profiler— a 23-item measure developed by Butler & Kern (2016) based on Martin Seligman’s PERMA model of flourishing (Positive Emotion, Engagement, Relationships, Meaning, Accomplishment). We use the official item wording and the authors’ published scoring method — each domain reported separately, never collapsed into one composite by default.

How Saati Insights are built

Our Sleep, Study Habits, Focus, and Stress questions are written in-house, informed by general, well-established research areas (sleep hygiene, self-regulated learning, attentional control, and perceived stress). They are deliberately written in plain, non-clinical language, scored as a simple average converted to a 0–100 scale, and never described using words like “validated,” “clinical,” or “diagnostic.”

What your report includes

Alongside your scores, Saati generates a short AI-written narrative summary to help make sense of the numbers together. The AI never computes a score itself — it only writes about scores that have already been calculated deterministically — and it’s instructed to never use diagnostic language, never position itself as crisis support, and to always keep validated and proprietary results clearly distinguished in its own wording. If it’s ever unavailable, you still get a complete report: your full report never depends on the AI.