Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/tjmahr/motorspeech2024

Poster for our presentation at the 2022 Motor Speech conference in San Diego, CA.
https://github.com/tjmahr/motorspeech2024

Last synced: about 1 month ago
JSON representation

Poster for our presentation at the 2022 Motor Speech conference in San Diego, CA.

Awesome Lists containing this project

README 

        
# Motorspeech2024



**Speaking rate and intelligibility in children’s speech**



Poster for our presentation at the 2022 Motor Speech conference in San

Diego, CA.



## Abstract



Prior work has documented how children’s intelligibility and

articulation rate increased over the childhood years. We asked whether

speaking rate influenced intelligibility in a sample of 538 typically

developing children between ages 2;6 and 9;11 (years;months). Because

utterance length can influence intelligibility (more words providing

additional sentence context) and rate (more words requiring more

efficient speech in a single fluent utterance), we stratified rate and

intelligibility scores by utterance length. We found a small negative

effect of speaking rate on intelligibility at younger ages for longer

utterance lengths. For example, at age 4;0, an increase in rate of .5

syllables per second predicted a corresponding decrease in

intelligibility of 3–5 percentage points on 5–7-word utterances. For

shorter items or for children older than age 5, this rate effect is

negligible. Implications will be discussed.



## Slide of the week



(A modified version of the primary figure from the poster was used for

the UW–Madison Waisman Center slide of the week.)



**Expected intelligibility decreases with speaking rate for younger

children and longer utterances**



![four panel figure described below][sotw]



*Intelligibility as a function of speaking rate and age in different

utterance lengths.* Typically developing children produced 4–7-word

utterances in a repetition task. We measured their speaking rate in

syllables per second, and their intelligibility as the proportion of

words correctly transcribed by two listeners. Points represent observed

data with 1 point per child per panel. Lines represent the expected

marginal mean for the central age in each age bin (i.e., age 3;0 for

the 2;6–3;5 bin). Within each panel, we see the lines stacked in the

expected order: Children become more intelligible with age. The vertical

black line on each line is the median speaking rate for children in the

age bin, and the thicker region of the line marks the 25%–75% percentile

range. Within each panel, we see the medians move from left to right

with age: Children's habitual speaking rate increases with age. All the

lines follow a flat or slightly negative slope, indicating that

increases in speaking rate predict decreases in intelligibility. This

effect is small to negligible in the three oldest age bins, but for

younger children and for longer utterances, the trend suggests a

speed-accuracy tradeoff for habitual speech production.



## Disclosures



Tristan Mahr worked on this project while employed by the UW–Madison

Waisman Center. He has no relevant nonfinancial relationships to

disclose.



Katherine Hustad received a research grant from the NIH to collect these

data. She has no relevant nonfinancial relationships to disclose.



## License



GPL-3 license intended for source-code (.R) files. Other creative

material (words and images) is CC BY licensed:

.



[sotw]: ./slide-of-the-week/2024-03-mahr-hustad-slide-of-week-1200.png "Expected intelligibility decreases with speaking rate for younger children and longer utterances"