An Introduction to Bayesian Multilevel Models Using brms: A Case Study of Gender Effects on Vowel Variability in Standard Indonesian

Nalborczyk, Ladislas; Batailler, Cédric; Lœvenbruck, Hélène; Vilain, Anne; Bürkner, Paul - Christian

doi:10.1044/2018_jslhr-s-18-0006

Cited by 133 publications

(81 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…An introduction to Bayesian statistics is outside the scope of this paper. However, the interested reader is referred to [76] for an introduction to Bayesian multilevel modelling using the brms package.…”

Section: Discussionmentioning

confidence: 99%

Can we decode phonetic features in inner speech using surface electromyography?

et al. 2020

Self Cite

View full text Add to dashboard Cite

Although having a long history of scrutiny in experimental psychology, it is still controversial whether wilful inner speech (covert speech) production is accompanied by specific activity in speech muscles. We present the results of a preregistered experiment looking at the electromyographic correlates of both overt speech and inner speech production of two phonetic classes of nonwords. An automatic classification approach was undertaken to discriminate between two articulatory features contained in nonwords uttered in both overt and covert speech. Although this approach led to reasonable accuracy rates during overt speech production, it failed to discriminate inner speech phonetic content based on surface electromyography signals. However, exploratory analyses conducted at the individual level revealed that it seemed possible to distinguish between rounded and spread nonwords covertly produced, in two participants. We discuss these results in relation to the existing literature and suggest alternative ways of testing the engagement of the speech motor system during wilful inner speech production.

show abstract

Section: Discussionmentioning

confidence: 99%

Can we decode phonetic features in inner speech using surface electromyography?

et al. 2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…In addition to the frequentist LME models, we also fit parallel Bayesian multilevel models using the package brms (Bürkner, 2017(Bürkner, , 2018. These models allowed us to quantify how much our data supports the null or the alternative hypothesis (see Nalborczyk, Batailler, Loe venbruck, Vilain, & Bürkner, 2019;Nicenboim & Vasishth, 2016;Vasishth, Nicenboim, Beckman, Li, & Kong, 2018 for descriptions of Bayesian multilevel models in the context of psycholinguistic research). We chose an ex-gaussian distribution model because it provides a considerably better fit for reaction time data which is typically (and also clearly in the present studies) right-skewed (Lindelø v, 2020; Rousselet & Wilcox, 2019).…”

Section: Methodsmentioning

confidence: 99%

Processing symbolic magnitude information conveyed by number words and by scalar adjectives

Kochari¹

2020

Preprint

View full text Add to dashboard Cite

Humans not only process and compare magnitude information such as size, duration, and number perceptually, but they also communicate about these properties using language. In this respect, a relevant class of lexical items are so-called scalar adjectives like ‘big’, ‘long’, ‘loud’, etc. which refer to magnitude information. It has been proposed that humans use an amodal and abstract representation format shared by different dimensions, called the generalized magnitude system (GMS). In this paper, we test the hypothesis that scalar adjectives are symbolic references to GMS representations, and, therefore, GMS gets involved in processing their meaning. Previously, a parallel hypothesis on the relation between number symbols and GMS representations has been tested with the size congruity paradigm. The results of these experiments showed interference between the processing of number symbols and the processing of physical (font-) size. In the first three experiments of the present study (total N=150), we used the size congruity paradigm and the same/different task to look at the potential interaction between physical size magnitude and numerical magnitude expressed by number words. In the subsequent three experiments (total N=149), we looked at a parallel potential interaction between physical size magnitude and scalar adjective meaning. In the size congruity paradigm we observed interference between the processing of the numerical value of number words and the meaning of scalar adjectives, on the one hand, and physical (font-) size, on the other had, when participants had to judge the number words or the adjectives (while ignoring physical size). No interference was obtained for the reverse situation, i.e. when participants judged the physical font size (while ignoring numerical value or meaning). The results of the same/different task for both number words and scalar adjectives strongly suggested that the interference that was observed in the size congruity paradigm was likely due to a response conflict at the decision stage of processing rather than due to the recruitment of GMS representations. Taken together, it can be concluded that the size congruity paradigm does not provide evidence in support the hypothesis that GMS representations are used in the processing of number words or scalar adjectives. Nonetheless, the hypothesis we put forward about scalar adjectives is still is a promising potential line of research. We make a number of suggestions for how this hypothesis can be explored in future studies.

show abstract

“…This prior was used for the effect of each predictor in each model. The models were run in the R package brms (Bürkner, 2017(Bürkner, , 2018, which is a high-level interface for Stan -an open source platform for full Bayesian statistical inference with MCMC sampling (Carpenter et al, 2017). For the models which used random effects, the brms default prior for random effects was used; all priors are detailed in Supplementary Materials.…”

Section: Quantitative Data Analysismentioning

confidence: 99%