Economic and Social Research Council – New Investigator:

• “Can you use it in a sentence?: Establishing how word-production difficulties shape text formation” (2023-2025) £304,812.00

US National Science Foundation (NTU: Torrance; External: Iowa State University, TU Eindhoven):

• “SourceWrite: Real-time, biometric, intention-informed scaffolding of source-based writing processes” (2024-2027) £669,647.00
• “ProWrite: Biometric feedback for improving college students’ writing processes.” (2020-2024) £592,695.00

The Leverhulme Trust (NTU: Jones, Torrance)

• “The time-course of statistical learning” (2024-2027) £340,090.00

A few unfunded ongoing projects, e.g.:

• Psycholinguistics of Tagalog (e.g. Garcia, Roeser, and Kidd 2023)
• External: Garcia (ZAS Berlin), Kidd (The Australian National University)

• Real-time automatised feedback for writers producing texts (essays) from sources.
• The model captures the reader’s current mental model of the source text, on the basis of eye-movement data.
• In combination with generative AI (semantic mapping between source texts and produced text), this allows the system to provide feedback to the writer of the form.

“We see that you’re writing about X. This is because you read Y, which is also similar to Z which you’ve not read. Also, you’ve missed the main point in Source 2.”

• This model directly informed by methods and findings from existing more theoretical work (Roeser et al. 2025).