Research compendium for our article:

Mahr, T., McMillan, B. T. M., Saffran, J. R., Ellis Weismer, S., & Edwards, J. (2015). Anticipatory coarticulation facilitates word recognition in toddlers. Cognition, 142, 345–350. 10.1016/j.cognition.2015.05.009

Repository overview

• data contains the eye-tracking data, child demographics and model-ready aggregated eye-tracking data.
  • data/README provides a "codebook" for the data in each file.
• R contains scripts that aggregate the eye-tracking data, fit lme4 models, as well as utility functions for formatting tables, numbers and other R objects.
• reports holds RMarkdown files that are used to produce formatted tables, formatted citations, and number-heavy paragraphs which appear in the article.
• plots contains the figures for the article.
• phonetics houses Praat scripts.
• stimuli contains sounds and images used in the experiment.
• extras are like the "bonus features" from the project.
• packrat is helper directory for managing a custom R package library.

The shell script build.sh runs all the R scripts to aggregate the data, fit the models, and generate HTML output from the RMarkdown files.

Tools

This repository works best with a current version of RStudio. Current (2015+) versions of RStudio include pandoc, and it is possible to render RMarkdown files or run the build.sh script with a single click. Use New Project > Version Control > Git > ... to clone this repository as an RStudio project. Once it's cloned, the packrat bootstrapper will download the packages and recreate the package library used for this repository. This process takes a while -- feel free to reread the article! (If that process fails, instead just manually install the packages listed in libraries.R.) Once the bootstrapper finishes, click Build All in the Build tab to re-run all the analyses and re-render the RMarkdown documents.

License

The GPL-2 license applies to the R code I have written in the .R and .Rmd files and to the Praat scripting code in the .praat files. The data, collected at the University of Wisconsin–Madison, still belong to the university.

Citation

Here is the BibTeX entry for the article:

@article {Mahr2015,
title    = {Anticipatory coarticulation facilitates word recognition 
            in toddlers},
author   = {Mahr, Tristan and McMillan, Brianna T. M. and Saffran, Jenny R. 
            and {Ellis Weismer}, Susan and Edwards, Jan},
year     = {2015},
journal  = {Cognition},
pages    = {345--350},
volume   = {142},
doi      = {10.1016/j.cognition.2015.05.009},
url      = {http://doi.org/10.1016/j.cognition.2015.05.009},
abstract = {Children learn from their environments and their caregivers. 
            To capitalize on learning opportunities, young children have to 
            recognize familiar words efficiently by integrating contextual 
            cues across word boundaries. Previous research has shown that 
            adults can use phonetic cues from anticipatory coarticulation 
            during word recognition. We asked whether 18--24 month-olds (n=29) 
            used coarticulatory cues on the word "the" when recognizing the 
            following noun. We performed a looking-while-listening eyetracking 
            experiment to examine word recognition in neutral versus 
            facilitating coarticulatory conditions. Participants looked to the 
            target image significantly sooner when the determiner contained 
            facilitating coarticulatory cues. These results provide the first 
            evidence that novice word-learners can take advantage of 
            anticipatory sub-phonemic cues during word recognition.}
}