Vocabulary Size Influences Spontaneous Speech in Native Language Users: Validating the Use of Automatic Speech Recognition in Individual Differences Research

Abstract

Previous research has shown that vocabulary size affects performance on laboratory word production tasks. Individuals who know many words show faster lexical access and retrieve more words belonging to pre-specified categories than individuals who know fewer words. The present study examined the relationship between receptive vocabulary size and speaking skills as assessed in a natural sentence production task. We asked whether measures derived from spontaneous responses to everyday questions correlate with the size of participants’ vocabulary. Moreover, we assessed the suitability of automatic speech recognition (ASR) for the analysis of participants’ responses in complex language production data. We found that vocabulary size predicted indices of spontaneous speech: individuals with a larger vocabulary produced more words and had a higher speech-silence ratio compared to individuals with a smaller vocabulary. Importantly, these relationships were reliably identified using manual and automated transcription methods. Taken together, our results suggest that spontaneous speech elicitation is a useful method to investigate natural language production and that automatic speech recognition can alleviate the burden of labor-intensive speech transcription.

A behavioural dataset for studying individual differences in language skills

“…Such tasks have been used widely in the neuropsychological and aging literature and allow for the analysis of various properties of spontaneous speech including speech rate, fluency (frequency and duration of silent and filled pauses), type-token ratio of words, average word frequency, utterance length, syntactic and conceptual richness. Moreover, in a recent pilot study, we have shown that measures extracted from spontaneous speech correlated positively with participants' receptive vocabulary size 47 .…”

IDLaS-NL – A platform for running customized studies on individual differences in Dutch language skills via the internet

“…Funded by the 'Language in Interaction consortium' (https://www.languageininteraction.nl/), IDLaS-NL was developed by researchers based at the Max Planck Institute for Psycholinguistics and at the Donders Institute for Brain, Cognition and Behavior between 2017 and 2022. To construct the battery, we conducted several pilot studies featuring parts of the tests (Brysbaert et al, 2021;Hintz et al, 2020a;Kapteijns & Hintz, 2021;Jongman et al, 2021) as well as one large pilot study that involved testing 112 participants twice on all battery tests, with approximately one month's time in between, for assessing test-retest reliability (Hintz et al, 2020b;Hintz et al, 2022). Our working model (McQueen & Meyer, 2019) assumed that variability in linguistic processing skills (the skills at stake here are word-and sentence-level processing) is influenced by variability in linguistic experience (leading to linguistic knowledge) and general cognitive skills.…”

The Dutch Auditory & Image Vocabulary Test (DAIVT): A New Dutch Receptive Vocabulary Test for Students

“…Our primary incentive to develop a new test was the lack of a freely available alternative for the PPVT to be used in research with young adult advanced users of Dutch (typically university students). The PPVT-III-NL is a popular instrument in research to assess Dutch receptive vocabulary size (e.g., Benard et al, 2014 ; Jongman, Khoe, & Hintz, 2020 ; Mainz et al, 2017 ; Simonis, Van der Linden, Galand, Hiligsmann, & Szmalec, 2020 ) and existing research suggests that it performs well for this population. Mainz et al ( 2017 ), for instance, found that the students’ PPVT-scores were successful predictors for performance (speed and accuracy) on a lexical processing task, with higher vocabulary scores being associated with better performance.…”

“…The PPVT test scores must be calculated manually, which is also time intensive. Some researchers avoided these problems by making their own online version of the PPVT-III-NL ( Jongman et al, 2020 ; Mainz et al, 2017 ), but because the test is copyright protected it cannot be shared with other researchers.…”