Jason H. Davidow

Systematic studies of modified vocalization: the effect of speech rate on speech production measures during metronome-paced speech in persons who stutter

BACKGROUND Metronome-paced speech results in the elimination, or substantial reduction, of stuttering moments. The cause of fluency during this fluency-inducing condition is unknown. Several investigations have reported changes in speech pattern characteristics from a control condition to a metronome-paced speech condition, but failure to control speech rate between conditions limits our ability to determine if the changes were necessary for fluency. AIMS This study examined the effect of speech rate on several speech production variables during one-syllable-per-beat metronomic speech in order to determine changes that may be important for fluency during this fluency-inducing condition. METHODS & PROCEDURES Thirteen persons who stutter (PWS), aged 18-62 years, completed a series of speaking tasks. Several speech production variables were compared between conditions produced at different metronome beat rates, and between a control condition and a metronome-paced speech condition produced at a rate equal to the control condition. OUTCOMES & RESULTS Vowel duration, voice onset time, pressure rise time and phonated intervals were significantly impacted by metronome beat rate. Voice onset time and the percentage of short (30-100 ms) phonated intervals significantly decreased from the control condition to the equivalent rate metronome-paced speech condition. CONCLUSIONS & IMPLICATIONS A reduction in the percentage of short phonated intervals may be important for fluency during syllable-based metronome-paced speech for PWS. Future studies should continue examining the necessity of this reduction. In addition, speech rate must be controlled in future fluency-inducing condition studies, including neuroimaging investigations, in order for this research to make a substantial contribution to finding the fluency-inducing mechanism of fluency-inducing conditions.

Strategies for Teachers to Manage Stuttering in the Classroom: A Call for Research

Purpose This clinical focus article highlights the need for future research involving ways to assist children who stutter in the classroom. Method The 4 most commonly recommended strategies for teachers were found via searches of electronic databases and personal libraries of the authors. The peer-reviewed evidence for each recommendation was subsequently located and detailed. Results There are varying amounts of evidence for the 4 recommended teacher strategies outside of the classroom, but there are no data for 2 of the strategies, and minimal data for the others, in a classroom setting. That is, there is virtually no evidence regarding whether or not the actions put forth influence, for example, stuttering frequency, stuttering severity, participation, or the social, emotional, and cognitive components of stuttering in the classroom. Conclusion There is a need for researchers and speech-language pathologists in the schools to study the outcomes of teacher strategies in the classroom for children who stutter.

Intrajudge and Interjudge Reliability of the Stuttering Severity Instrument–Fourth Edition

Purpose The Stuttering Severity Instrument (SSI) is a tool used to measure the severity of stuttering. Previous versions of the instrument have known limitations (e.g., Lewis, 1995). The present study examined the intra- and interjudge reliability of the newest version, the Stuttering Severity Instrument-Fourth Edition (SSI-4) (Riley, 2009). Method Twelve judges who were trained on the SSI-4 protocol participated. Judges collected SSI-4 data while viewing 4 videos of adults who stutter at Time 1 and 4 weeks later at Time 2. Data were analyzed for intra- and interjudge reliability of the SSI-4 subscores (for Frequency, Duration, and Physical Concomitants), total score, and final severity rating. Results Intra- and interjudge reliability across the subscores and total score concurred with the manuals reported reliability when reliability was calculated using the methods described in the manual. New calculations of judge agreement produced different values from those in the manual-for the 3 subscores, total score, and final severity rating-and provided data absent from the manual. Conclusions Clinicians and researchers who use the SSI-4 should carefully consider the limitations of the instrument. Investigation into the multitasking demands of the instrument may provide information on whether separating the collection of data for specific variables will improve intra- and interjudge reliability of those variables.

Stuttering Frequency, Speech Rate, Speech Naturalness, and Speech Effort During the Production of Voluntary Stuttering

Purpose: Voluntary stuttering techniques involve persons who stutter purposefully interjecting disfluencies into their speech. Little research has been conducted on the impact of these techniques on the speech pattern of persons who stutter. The present study examined whether changes in the frequency of voluntary stuttering accompanied changes in stuttering frequency, articulation rate, speech naturalness, and speech effort. Method: In total, 12 persons who stutter aged 16–34 years participated. Participants read four 300-syllable passages during a control condition, and three voluntary stuttering conditions that involved attempting to produce purposeful, tension-free repetitions of initial sounds or syllables of a word for two or more repetitions (i.e., bouncing). The three voluntary stuttering conditions included bouncing on 5%, 10%, and 15% of syllables read. Friedman tests and follow-up Wilcoxon signed ranks tests were conducted for the statistical analyses. Results: Stuttering frequency, articulation rate, and speech naturalness were significantly different between the voluntary stuttering conditions. Speech effort did not differ between the voluntary stuttering conditions. Stuttering frequency was significantly lower during the three voluntary stuttering conditions compared to the control condition, and speech effort was significantly lower during two of the three voluntary stuttering conditions compared to the control condition. Conclusions: Due to changes in articulation rate across the voluntary stuttering conditions, it is difficult to conclude, as has been suggested previously, that voluntary stuttering is the reason for stuttering reductions found when using voluntary stuttering techniques. Additionally, future investigations should examine different types of voluntary stuttering over an extended period of time to determine their impact on stuttering frequency, speech rate, speech naturalness, and speech effort.

Comments on Brown et al. (2016)

Brown et al. (2016) recently reported the results of a study with seven adult persons who stutter (PWS) who completed a “prolonged speech” – based treatment. The study focused on the speakers’ phonatory behavior before and after the establishment phase (specifically, on the final day of establishment) of their treatment program. The analysis most relevant to the present discussion was an examination of the frequency of phonation intervals (PIs; basically, an estimate of the duration of vocal fold vibration) between 31 and 200 ms in duration in segments of stutterfree speech. For this assessment, Brown et al. separated the PIs into bins, resulting in four separate bins for the shortest PIs: 31–50 ms, 51–100 ms, 101–150 ms, and 151–200 ms. Perhaps the most important conclusion presented by Brown et al. (2016) was that stuttering can be reduced to very low levels following a prolonged speech treatment without reductions in the frequency of short PIs. This conclusion is important because it has been argued previously that reductions in the absolute or relative frequency of short PIs might be functionally associated with reduced stuttering in PWS during fluency-inducing conditions, including treatments that rely on modifying phonation (Ingham et al., 2001). Brown et al. reported that 3 of their 7 participants failed to show a reduction of 50% or more in the frequency of short PIs in any of the four bins outlined above, during post-establishment phase improved fluency; one participant actually showed a 50% or more increase in one bin. In consequence, the investigators concluded that a reduction in the frequency of short PIs “does not appear necessary for this reduction in stuttering to occur” (Brown et al., 2016, p. 13). There are, however, some important reasons to question whether the presented data can sustain either this specific conclusion or its larger implications. Several of these reasons are addressed below. First, the use of the fixed 50-ms bins (and the shortest 20-ms bin) may not be the best way to examine this issue. PI duration is a continuous variable; bins are used as an artificial means of creating a categorical variable for analysis. Any result, therefore, “might be an artifact of the cutoff time boundaries for each bin” (Davidow et al., 2009Davidow, Bothe, Andreatta, & Ye, 2009, p. 199). This is compounded by the possibility that the use of too many bins can prevent the identification of changes that did occur and that would have been identified across a larger window of analysis, as well as by the possibility that the specific PI durations that need to be reduced may differ across speakers. Within the Modifying Phonation Interval (MPI) treatment program (Ingham& Student, 2014), for example, speakers are assigned individualized “Target PI” ranges based on their own baseline PI distribution. This Target PI range is usually between 30 ms and 200 ms, as Brown et al. noted, but it cannot be assumed to be 30–200 ms and it cannot be assumed to be any of the 50-ms bins that Brown et al. investigated. Second, the main conclusion of the Brown et al. (2016) study is based on whether or not a 50% reduction in PIs occurred in any one of four PI bins. As discussed previously (Davidow et al., 2009; Ingham et al., 2001), “each person who stutters may benefit from a different target range reduction percentage” (Davidow et al., 2009, p. 202). That is, the amount of reduction needed to induce improved fluency may be less than 50% for some PWS. Of course, it would also be interesting to learn if the participants who did not show a 50% reduction in the frequency of short PIs would have greater reductions in stuttering with a 50% reduction. Third, and very much related, neither post-treatment PI distribution data nor comparative (pre-post) PI distribution data were provided by Brown et al. (2016), and therefore it was not possible for readers to determine whether or how changes occurred in PI durations or distributions. Brown et al. did report, in their Table 4, which of the four bins below 201 ms “showed changes of 50% or greater” (Brown et al., 2016, p. 18) for each speaker, for one comparison between segments of stutterfree pretreatment speech and segments of stutterfree posttreatment speech. However, post-treatment PI distributions would have allowed the reader to extract more information, including the magnitude of decrease or increase in each bin. Fourth, it is important to recognize that the frequency of short PIs could have been reduced during the initial stages of treatment, and could have been functionally important to treatment effects, even if the changes were no longer evident in the posttreatment data gathered by Brown et al. (2016). Evidence for this possibility stems from Davidow et al.’s (2009) demonstration that the percentage of short PIs was reduced shortly after learning how to produce prolonged speech at rates below 180 syllables per minute (SPM). In fact, all of the participants in Davidow et al.’s study had a reduction in at least 1 of 3 bins (30–50 ms, 51–100 ms, and 101–150 ms). In other words, the improved speech shown by the Brown et al. participants at the post-establishment measurement point could very possibly have been the product of a treatment that had included a reduction in the frequency of short PIs during earlier stages of training. Brown et al. themselves highlighted this important idea by presenting the following quote from the Davidow et al.