Jasbi, Jaggi, Frank (2018): Conceptual and prosodic cues assist disjunction acquisition
Structure of the repo:
1_raw_data: contains the raw data used in the study. Files starting with "Providence" belong to our study 2: the annotation of disjunction in child directed speech in the Providence corpus of CHILDES. Other files belong to study 1. They record what we extracted from childes-db.
2_processed_data contains the dataframes used for study 1 plots. We obtain them by normalizing the frequency of the connectives by either the total number of words spoken by a speaker at a particular age (raw_normalized.csv) or the number of words in a particular speech act (bySpeechAct_normalized.csv).
figs contains the figures in the paper.
To reproduce, please follow these steps:
cogsci2016package. You can find the pakcage and the instructions here.
childesrpackage. The github repo and instructions are here.
Make sure the following R packages are installed:
jpeg(0.1.8). If you want the packages installed at once use the following code:
install.packages(c("png", "grid", "ggplot2", "xtable", "knitr", "tidyverse", "ggthemes", "lubridate", "magrittr", "lme4", "lmerTest", "forcats", "bootstrap", "jpeg")).
Make sure you have an updated verison of LaTeX. You can install TeX by visiting the LaTeX Project website.
Clone/download the repo here by clicking on the green button "Clone or download" on the right top of this page.
On your computer, open the main Rmarkdown file (JasbiJaggiFrank_cogsci2018.Rmd) and press Knit.
Hopefully the Rmd file renders with no error and you get the pdf file (JasbiJaggiFrank_cogsci2018.pdf). If you see any errors/problems while reproducing our work or if you have any questions please do not hesitate to contact the first author.