Peekbank: An open, large-scale repository for developmental eye-tracking data of children’s word recognition

Abstract

Author NoteAcknowledgements. We would like to thank the labs and researchers that have made their data publicly available in the database. For further information about contributions, see https://langcog.github.io/peekbank-website/docs/contributors/. This work was supported in part by grants from the National Institutes of Health awarded to VM (R01 HD092343, 2R01 HD069150).Open Practices Statement. All code for reproducing the paper is available at https://github.com/langcog/peekbank-paper. Raw and standardized datasets are available on the Peekbank OSF repository (https://osf.io/pr6wu/) and can be accessed using the peekbankr R package (https://github.com/langcog/peekbankr).CRediT author statement. Outside of the position of the first and the last author, authorship position was determined by sorting authors' last names in reverse alphabetical order. An overview of authorship contributions following the CRediT taxonomy can be viewed here: https://docs.google.com/spreadsheets/d/e/2PACX-1vRD-LJD_dTAQaAynyBl wXvGpfAVzP-3Pi6JTDoG15m3PYZe0c44Y12U2a_hwdmhIstpjyigG2o3na4y/pubhtml.

The more they hear the more they learn? Using data from bilinguals to test models of early lexical development

“…First, future studies could test bilinguals using both types of measures, such that models can be fitted on data from the same group of infants (for an example of a related approach, see Fernald et al, 2012). Emerging resources such as Peekbank, a repository of looking-while-listening data with associated measures such as CDI, might be especially valuable for gathering large and diverse datasets (Zettersten et al, 2022). Second, to address differences related to binary measures from parent report on the one hand and gradient measures from behaviour on the other hand, gradient parent report measures could be developed that assess the strength and/or consistency of children's lexical comprehension and/or production.…”

Daylong egocentric recordings in small- and large-scale language communities: A practical introduction

“…It also fortifies the existing network of data contributors and re-users in the language sciences, improves the quality of existing datasets, and thus helps ensure that these valuable resources will be available to students and scientists long term. As a field, developmental researchers have called for and demonstrated an enormously successful history of data sharing (Adolph, Gilmore, et al, 2012;Gennetian et al, 2020;Gilmore, 2022;Kosie & Lew-Williams, 2022), thanks to the generosity of participating families, contributing researchers, and the scientific leaders who implement and maintain digital sharing infrastructures (Frank et al, 2017;Gilmore et al, 2016;MacWhinney, 2000;VanDam et al, 2016;Zettersten et al, 2023).…”

Valid points and looks: Reliability and validity go hand‐in‐hand when improving infant methods

“…Improved modelling techniques have begun to parse the components that shape infants' looking behaviour in a more fine‐grained manner (McMurray, 2020; Piantadosi et al, 2014; Poli et al, 2020). Efforts to combine pre‐existing data into large‐scale databases give us sufficient power to systematize our understanding of the measurement properties of looking‐time data (Zettersten et al, 2021; Zettersten, Yurovsky, et al, 2022). Finally, large‐scale multi‐site efforts such as the ongoing ManyBabies5 project can help us understand the degree to which variance in looking time is predicted by properties of participants (e.g., age) and properties of the task design (e.g., familiarization time).…”