What is possible with speecollectr

The need for remote speech data collection during the Covid-19 pandemic prompted the development of the open-source software speechcollectr, an R package for building speech experiment interfaces. This package relies on Shiny, an R package for web development. Experiments built with speechcollectr can be run from a local machine or distributed over the Internet for web-based protocols. Designed primarily for researchers familiar with R, this tutorial will demonstrate some foundational concepts of reactive programming in R. This type of programming allows researchers to construct interfaces that respond to participant inputs. Such interfaces include surveys or audio recording and playback interfaces, all of which may be created and displayed to participants with speechcollectr functions. A key feature of the package is its suite of functions for recording uncompressed WAV files from the participant's browser. Recordings created with speechcollectr are uploaded automatically to the experimenter's server and may be subjected to the automated acoustic evaluation performed by speechcollectr's evalWav function. This tutorial will include a brief walkthrough of a speech production task to demonstrate the speechcollectr protocol for audio recording and, more broadly, to show how speech data may be collected using R.