Education Data Science

Linguistic Indicators of the Growth Mindset

Project Year
2023
Abstract

We analyze students’ essay responses and post-intervention outcomes(n= 4059) to find how language might correlate with stress levels. We use data from the synergistic mindset intervention that contains students’ essay responses and self-reported stress appraisals after a self-administered online training module intervention. With the aid of word count based approaches, Latent Dirichlet Allocation, and word embedding based methods, we analyse the relationship between linguistic patterns and topics in student responses with their demographics and stress levels. Results show that students that cope well with stress use a greater use of negative over positive emotions , past and present tense over future tense words, higher cognitive processing words, and more action over bodily sensation words in comparison to those that cope poorly. We discuss our findings in relation to previous linguistic studies of mindfulness and stress.

EDS Students