SSI Score
Scoring Pattern
SSI = average(8 feature scores)→structural preachiness measure (0-5 scale)
Definition
The Structural Sermonic Index score — computed as the simple average of all eight SSI feature scores (imperative density, second person address, absolute claims, moral framing, rhetorical questions, emotional escalation, binary framing, and call to action). Each feature is scored 0-5, and the SSI averages them into a single 0-5 measure of structural preachiness. The SSI captures how the text is organized: its use of commands, direct address, absolutism, moral framing, rhetorical questions, escalation, binary thinking, and exhortation. A high SSI means the text is structurally built like a sermon, regardless of its specific vocabulary.
Examples
Example 1
SSI 4.2: imperative density 5, second person 4, absolute claims 4, moral framing 5, rhetorical questions 3, emotional escalation 4, binary framing 4, call to action 5.
A high SSI score showing heavy structural preachiness across all features — the text is built like a sermon at every level of organization.
Example 2
SSI 1.3: imperative density 2, second person 1, absolute claims 1, moral framing 2, rhetorical questions 1, emotional escalation 1, binary framing 1, call to action 2.
A low SSI showing mild advisory structure — traces of prescriptive patterns without sustained sermonic organization.
Example 3
SSI 2.8: imperative density 4, second person 3, absolute claims 2, moral framing 4, rhetorical questions 2, emotional escalation 3, binary framing 2, call to action 3.
A moderate SSI with uneven feature distribution — high on imperatives and moral framing but lower on absolutism and binary framing, suggesting passionate advocacy rather than full-spectrum preaching.
AI Detection Note
AI text typically scores SSI 1.5-2.5 — moderate structural preachiness from its default advisory tone. Human text has much wider SSI variance: genuinely neutral writing scores near 0, while human sermons routinely score 3.5-5.0. The SSI range of a text is itself diagnostic: narrow range (all features similar) suggests AI; wide range (some features very high, others very low) suggests human authorship.
See how your writing scores on the Sermonic-Preachy Index
Analyze Your Text