Multimodal analysis (MMA) is increasingly used to identify autistic behaviours, supporting timely access to educational and therapeutic support. These tools rely on behavioural markers — eye gaze, prosody and gestures — that are culturally variable in expression and interpretation. Because most MMA methods and datasets originate in Western clinical settings, they risk systematically misidentifying behaviour in non-Western and Indigenous families. In Aotearoa, Māori autism isreflected as takiwātanga, yet tools built on Western deficit norms may not recognise takiwātanga, risking inequitable support for tamariki and whānau Māori.
Our final goal is to enable culturally responsive tools that help parents recognise their child's unique behaviours early. To achieve this, it is important to answer the question: do current MMA tools hold cultural considerations and are validated beyond Western settings? Building tools before mapping this evidence risks encoding the very cultural bias the work seeks to remove.
This scoping review therefore maps (a) the behavioural markers and modalities used by current MMA approaches to identify autistic children’s behaviour, (b) evidence about the cultural variability of those markers. It also examines whether MMA methods, datasets, or tools have been developed or validated across Western, non-Western, and Indigenous contexts, and identifies gaps limiting culturally responsive approaches.
Supervisors
Primary Supervisor: Claris Chung
Other Supervisor(s): Dean Sutherland
Application and funding
This is a UC Commitments Summer Scholarship research project. Emailed applications will not be considered. Find out more here: University of Canterbury Scholarship Portal - UC Commitments Summer Scholarship
Final date for receiving applications
25 September 2026
