This thesis introduces a new theoretical framework from which to view the factors that contribute to stuttering variability. The speech and monitoring interaction (SAMI) framework proposes that there are two systems that account for stuttering variability: the speech production system and the monitoring system. Each system has unique factors that modulate them. Within SAMI, the interaction of these two categories of factors is formalized in a mathematical equation. What is particularly novel about SAMI is the use of a mathematical equation to formalize the interaction between these systems and the specific proposal of the biological substrates of the monitoring system and its interaction with the speech system. The focus of this thesis is on the anticipation of stuttering, which is one of the factors from SAMI that modulates the monitoring system. The goals of the studies were to 1) characterize the degree to which people who stutter (PWS) anticipate stuttering and how accurately they can predict actual stuttering, 2) investigate the correlation between stuttering expectancy on words and the verbal response time to say those word in word naming tasks, and 3) make a qualitative comparison of the behavioral results and the results from a neural network model. Utilizing the SAMI framework it was hypothesized that stuttering expectancy would be positively correlated with the response time and the results from the simulations would qualitatively match the behavioral results.

The key finding was that across the group of PWS, there was positive correlation between stuttering expectancy scores and relative reaction times on those words. The degree to which stuttering expectancy was correlated with reaction time within subjects was positively correlated with stuttering severity. A qualitative comparison showed a good fit in between results of the simulations and the behavioral study. This is the first study to show that the expectation of stuttering has an effect on fluent speech production, providing evidence that the anticipation of stuttering is not only correlated with moments of stuttering but may also be a contributing factor to stuttering. The model provides a means of hypothesizing and testing specific neural substrates associated with anticipation of stuttering and its effects on the speech production process.