Language Experiences Overview: an Individualised Report Describing Children’s Multilingual Language Exposure

• In Event: Poster Session 01
  In Poster Session: PS 01 Topic - Methods, History, Theory

Abstract

Background. How do we characterize patterns of language exposure in meaningful ways? Singaporean families often involve grandparents in the care of young children, and one in five families employ a live-in foreign worker for domestic help (Singapore Ministry of Manpower, 2020). Hence, infants in Singapore typically hear two or more languages from each parent, two-to-three languages from their grandparents (Woon, 2018), and possibly more from a domestic helper. Given this complexity, existing models of balanced/unbalanced or dominant/non-dominant bilingualism are insufficient for describing the rich tapestry of multi-lingual experiences. In order to capture this variety, we created a flexible multilingual tool, the Language Experience Overview (LEO). This multivariate multilingual tool combines estimates of care time and language-use ratios with language profiles of each carer. Importantly, the LEO Report visualizes the results in parent-friendly feedback.
LEO Measurement Properties. Parents are asked to estimate care-time for all caregivers in a typical week, and to estimate their various language proficiencies on a scale from 1 (only know a few words) to 7 (Native). Parents also estimate what proportion of their time each carer uses different languages/dialects with the child (e.g. Mother: 100% English; Father: 50% English, 40% Mandarin Chinese, 10% Cantonese). Proportion of care-time is converted to a weighted average (e.g., 50:50), and the carer’s mix of languages is combined into a speaker-by-language matrix totalling 100% of documented language exposure. The matrix is used to derive a Composite Language Input Profile (CLIP) for the child: a proportional representation of the language mix of their input from carers (e.g., English: 75%, Mandarin Chinese 20%, 5% Cantonese).
LEO Report. The LEO Report consolidates a child’s multilingual input in a graphical format suitable for parents, educators and clinicians. The child’s CLIP is presented as a pie chart of their total language mix. Each carer’s contribution is shown, with relative size indicating contribution to the CLIP, along with their Language Fingerprint - a spider diagram showing estimated proficiency for each caregiver on four axes: understanding, speaking, reading and writing. This visualization, shown in Figure 1, allows a consolidated overview of how the language skills of each carer are contributing to the language exposure of this child, providing opportunities for reflection, insight and planning.
Technology Development. The LEO questions, and script for generating print-ready LEO Reports are an Open Access research tool suitable for adaptation into any combination of languages. (Archive: http://anonymouslink). We are currently creating a web-based interface for researchers, parents, and educators to enter their data and automatically generate LEO reports.
Current and Future Directions. Metrics from the LEO can be used as predicters of multilingual language development. In one ongoing RCT of parent-child language use, 170 parents of children under 3 have completed the LEO questions via a Qualtrics interface. After the study is complete (February 2021), parental feedback about the LEO Reports will be collated. This feedback will be critical for developing usage guides for parents on how they can use the LEO report for home language planning, and in consultation with educators and clinicians.

Authors

• Fei Ting Woon, Nanyang Technological University
  Presenting Author

• Tuan Anh Le, Nanyang Technological University
  Non-Presenting Author

• Shaza Binte Amran, Nanyang Technological University
  Non-Presenting Author

• Suzy J Styles, Nanyang Technological University, Singapore
  Non-Presenting Author