Using big data from long-form recordings to study development and optimize societal impact

Vocal input and output among infants in a multilingual context: Evidence from long‐form recordings in Vanuatu

“…This move is important as researchers in child development become increasingly aware of the low levels of linguistic and cultural diversity represented in our mainstream journals (Nielsen et al., 2017; Kidd & Garcia, 2022). Long‐form recordings hold many promises in this regard, including capturing children's spontaneous language use with minimal discomfort and intrusion for participants, particularly when paired with automatized algorithms like those used here, which were trained with multi‐cultural data allowing relatively high levels of accuracy even in multilingual corpora like ours (Cychosz & Cristia, 2023).…”

Long-form recordings in low- and middle-income countries: recommendations to achieve respectful research

“…Mehl et al, 2010). Cychosz and Cristia (2022) mention three such uses of child-centred recordings: (1) the detection of language disorders and potential treatment evaluations, (2) large-scale real-life interventions aimed at improving children's language and cognitive outcomes and (3) documenting under-documented languages.…”

“…One example of potential benefits related to long-form recordings is that the recordings can constitute a database to document under-represented languages or create other language resources based on how language is used in everyday activities and contexts (Cychosz and Cristia, 2022). Moreover, as with other types of language data, researchers with appropriate training and links to the software industry can use long-form recordings to build and train speech and language technology that the community considers useful.…”

“…They pose unusual ethical challenges and guidance has been developed for researchers using this technique (Cychosz et al, 2020). More recently Cychosz and Cristia (2022), discussed some additional ethical considerations. This paper builds on that foundation with consideration of long-form, child-centred recordings to help researchers working with big data approaches to collect data in LMICs ethically.…”

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

“…This chapter adds a unique perspective to the existing literature on daylong recording methods (e.g., Bergelson et al, 2019;Casillas, 2023;Cristia et al, 2021;Cychosz & Cristia, 2022;Gautheron et al, 2020;Montag, 2020): corpus building in the context of smaller-vs. larger-scale language communities. In the following two subsections, we illustrate how priorities and anticipated issues differed when we built corpora for two smaller-scale, subsistence communities (Tseltal and Yélî Dnye) compared to a larger-scale, urban/suburban U.S. English sample.…”