How to calculate percent syllables stuttered (%SS)
Percent syllables stuttered is the share of syllables in a speech sample that contain stuttering. It's the most common measure of how often someone stutters, and the arithmetic is simple. The hard parts are counting syllables in real time and getting a sample that reflects how the person really talks.
The formula
%SS = stuttered syllables ÷ total syllables × 100
In a 300-syllable conversation sample, 12 syllables were stuttered: 12 ÷ 300 × 100 = 4.0%SS.
%SS is usually reported to one decimal place, because the numbers are small: the difference between 2.5% and 3.4% matters.
What counts as a stuttered syllable
A syllable counts once if it contains any stuttering, however long the stutter lasts or however many times the sound is repeated. "B-b-b-ball" is one stuttered syllable.
Usually counted:
- Part-word repetitions: "b-b-ball," "ta-ta-table."
- Single-syllable word repetitions: "I-I-I want," which most protocols count.
- Prolongations: "sssssun."
- Blocks: a silent or tense stop before or inside a word.
Usually not counted as stuttered, though worth noting separately:
- Interjections: "um," "like."
- Revisions: "I went, I was going."
- Phrase repetitions: "I want, I want the red one."
These are typical disfluencies that every speaker produces. Protocols differ at the edges, so write down the rules you use and apply them the same way to every sample. If you're giving a published assessment, use that assessment's counting rules for it.
How big a sample?
Stuttering varies from moment to moment, so short samples bounce around. In a 100-syllable sample, each stuttered syllable moves the result by a full percentage point; in a 300-syllable sample, by a third of a point. Many clinicians aim for at least a few hundred syllables per sample.
Counting in real time
- Tap per syllable. Use one button or tally for each fluent syllable and another for each stuttered one. It takes practice, but the rhythm of tapping along with speech comes quickly.
- Record and count afterwards. This is slower but more accurate, and you can recount. Check your district's policy on recording students first.
- Check your own consistency. Count the same recorded sample twice, a week apart. If the two results differ by much, tighten your counting rules.
Sample in more than one situation
Many people who stutter are more fluent in the therapy room than in class, on the phone or with friends. A %SS from a structured conversation with you can be far lower than one from the classroom. Note the situation with every sample (reading aloud, monologue, conversation, classroom) and compare like with like.
%SS isn't the whole picture
%SS measures how often stuttering happens, not how much it affects the person. Two students at 3%SS can be in very different places: one avoids speaking in class, while the other talks freely. Pair %SS with the student's own view of their speech, how much they avoid words or situations, and how fully they take part in class. Many fluency goals today focus on communication and participation rather than on reducing stuttering itself.
Reporting it
Give the number, the sample type and the comparison: "In 6 conversation samples from January to March, Theo averaged 3.2% syllables stuttered, down from 7.5% at baseline." More examples are in the progress report comments guide.