Differentiating mild and recovered stutterers from nonstutterers

Abstract

In the present study, a diagnostic battery that included an interview scale (the Ease of Speech Production Scale, or ESPS) and two experimental perturbation tasks (a tongue twister task and a timed story construction task) was administered to 20 control speakers and 18 mild or recovered adults who stutter. The primary purpose of the study was to determine if this battery was successful at discriminating the mild and recovered adults who stutter from the control cases in this sample. Results indicated that the total score from the ESPS scale in combination with one of the perturbation task variables (off-target responses in a tongue twister task) was the best predictor of group membership in a preliminary predictive discriminant analysis, accurately classifying 94% of the persons who stutter and 95% of the control speakers. The ESPS total score, by itself, was almost as effective as the two-variable (best-fitting) model, accurately classifying 89% of the adults who stutter and 90% of the control cases. These findings suggest that a carefully constructed diagnostic interview has the potential to accurately discriminate mild and recovered stutterers who present few or no overt dysfluencies from nonstuttering control subjects. The ability to classify mild and recovered stutterers is particularly important for genetic and epidemiological research because these studies require the ability to assess the lifetime status of a large number of individuals, many of whom have recovered with or without formal treatment.