Throwback (2014): Help with the tricky business of diagnosing stuttering in preschoolers

Picture this: a parent asks you about her preschooler’s speech. She’s concerned that he might be stuttering. “How can I tell if it’s normal or not?” she asks.

Is your heart racing just thinking about this question? Preschool stuttering can be a big ol’ gray area; it’s hard to tell when normal disfluencies cross the line into developmental stuttering. Luckily, Tumanova et al. have some evidence-based answers for us.

The authors tested a big group of preschoolers (2.5- to 5-year-olds) and classified them as children who stutter or who do not stutter using a cutoff of 3% stuttered words from a 300-word speech sample and a score of 11 or greater on the Stuttering Severity Instrument–3. Then, they looked at how well the other factors they measured classified children into one group or the other. Remember that we evaluate diagnostic accuracy by considering sensitivity (how often the test correctly identifies a disorder) and specificity (how often the test correctly identifies typical development); both should be 80% or higher

Two measures stood out as having good diagnostic accuracy. First, percent total disfluencies—a combination of both stuttered disfluencies (e.g., prolongations, sound repetitions) and non-stuttered disfluencies (e.g., phrase repetitions, interjections)—had fair sensitivity (82%) and specificity (95%) at the cutoff of 8% disfluencies (the authors used a 300-word sample, but there is no standard recommended sample length; longer is better, and having two samples instead of one would also be better). Next, whether or not parents were concerned about their children’s stuttering had fair sensitivity (80%) and good specificity (92%) when compared to a cutoff of 3% stuttered disfluencies, “suggesting that the 3% criterion has a strong and clinically meaningful association with parental concern.” Also, boys had more non-stuttered (normal) disfluencies than girls; other factors (language ability, age) were related to how many normal disfluencies children produced, but the differences were small. 

So 3% stuttered disfluencies is still our go-to for diagnosing stuttering in preschoolers, but we can supplement that by looking at (1) percent non-stuttered (normal) disfluencies and (2) whether parents report concerns about their child’s stuttering.


Tumanova, V., Conture, E. G., Lambert, E. W., & Walden, T. A. (2014). Speech disfluencies of preschool-age children who do and do not stutter. Journal of Communication Disorders, 49, 25-41. doi: 10.1016/j.jcomdis.2014.01.003