Are there sex effects for speech intelligibility in American English? Examining the influence of talker, listener, and methodology

Abstract

Talker and listener sex in speech processing has been largely unknown and under-appreciated to this point, with many studies overlooking the possible influences. In the current study, the effects of both talker and listener sex on speech intelligibility were assessed. Different methodological approaches to measuring intelligibility (percent words correct vs. subjective rating scales) and collecting data (laboratoryvs.crowdsourcing)were alsoevaluated.Findings revealed that,regardless of methodology,thespokenproductions offemale talkers were overall more intelligible than the spoken productions of male talkers; however, substantial variability across talkers was observed. Findings also revealed that when data were collected in the lab, there was an interaction between talker and listener sex. This interaction between listener and talker sex was not observed when subjective ratings were crowdsourced from listener subjects across the USA via Amazon Mechanical Turk, although overall ratings remained similar. This possibly suggests that subjective intelligibility ratings may be vulnerable to bias, and such biases may be reduced by recruiting a more heterogeneous subject pool. Many studies in speech perception do not account for these talker, listener, and methodology effects. However, the present results suggest that researchers should carefully consider these effects when assessing speech intelligibility in different conditions, and when comparing findings across studies that have used different subject demographics and/or methodologies.

Perceiving gender while perceiving language: Integrating psycholinguistics and gender theory

“…For example, the degree of dispersion in individuals' vowel spaces—a reflection of the acoustic distinctiveness of individual vowels—predicted differences in intelligibility among the more‐intelligible talkers (a disproportionate number of whom were women), but only weakly predicted intelligibility across all 20 talkers. Similar findings are presented by Yoho et al (2019), who examined both intelligibility and subjective clarity ratings of sentences produced by 25 women and 25 men. They found large (~15%) differences between the average intelligibility of the women and men.…”

Single Word Intelligibility of Individuals with Parkinson’s Disease in Noise: Pre-Specified Secondary Outcome Variables from a Randomized Control Trial (RCT) Comparing Two Intensive Speech Treatments (LSVT LOUD vs. LSVT ARTIC)

“…That study [78] found no significant difference in single word intelligibility by gender in dysarthric participants with MS or PD nor in HCs. Females demonstrate a greater F0 variation when speaking [125], and it is known that a greater degree of F0 variation positively affects intelligibility [128][129][130]. However, a difference in F0 variability would more likely affect the intelligibility of sentences than the monosyllabic words used in the present study.…”

“…The relatively little research that has been devoted to this question in healthy speakers has come to conflicting conclusions. Some studies found female voices to be more intelligible than males [123][124][125], some found male talkers to be more intelligible than female talkers [126], and some found male and female voices to be equivocal in terms of intelligibility [127]. There is only one study that we are aware of that addresses this question in neurologically impaired speakers.…”

Spread the Word: Enhancing Replicability of Speech Research Through Stimulus Sharing

“…Identification rates for particular words depend on the talker's style of speaking. In addition to group differences that are known to affect speech intelligibility-speaker sex (Yoho et al, 2019) differences in language background (Bent & Bradlow, 2003;Clarke & Garrett, 2004) , and dialect variations (Clopper, 2021) -idiosyncrasies in speaking style that remain within the bounds of the accent and dialect with which the listener is most familiar can also affect speech identification (see McCloy et al, 2015) . Indeed, speech produced in a clear speaking style is identified more accurately than speech produced conversationally (Van Engen, 2017;Van Engen et al, 2012) , and speakers differ in the clarity of their speech (Smiljanić & Bradlow, 2005) .…”