Understanding Scores, Tiers, and Evidence Strength
How to read the numbers, tiers, and confidence labels in your report.
Your report uses a few consistent measures across every Dimension. Here's how to read them. (Every report also has a Methodology section at the bottom with these definitions.)
Dimension scores (out of 5)
Each Dimension gets an average score out of 5, based on what participants actually said. Behind each score, Savo breaks every conversation into small evidence units — snippets substantial enough to judge on their own — and rates each one from 1 to 5 against the dimension's definition. A participant's score is the average across their units; the dimension's score is the average across participants.
Tiers
- Strength — the cohort's average sits in the top band (about 4–5).
- Mixed — the average sits in the middle (about 3).
- Opportunity — the average sits in the low band (about 1–2).
- Evidence thin — tentative — fewer than about 20% of participants were scored on the dimension, so there isn't enough to band; you'll see this instead of a tier.
Confidence: how much evidence is behind a finding
- Strong — at least 1 in 5 of participants, with rich supporting quotes.
- Moderate — about 1 in 10, with consistent quotes.
- Limited — real signal, just thinner.
Score bands and evidence types
When you explore the evidence (see Exploring the Evidence), each quote is rated High (4–5), Moderate (3), or Low (1–2), and labeled by kind:
- Story — a specific, recalled experience with concrete detail (the strongest kind of evidence).
- Example — a concrete instance, without the full story.
- Claim — a general belief or opinion, without a specific instance.
When a participant isn't counted
To be scored on a Dimension, a participant needs enough evidence about it in their conversation. If they didn't talk about it enough, they're counted as “not heard” for that Dimension rather than guessed at — the report's Descriptives table shows the heard / not-heard counts per Dimension.