The Emotional Content of Children's Writing: A Data-Driven Approach

Speaker: Rainy Dong
Time: 17:30-18:30
Levett Room, Wolfson College & Online (Teams link)

My name is Rainy Dong and I'm a third-year DPhil in Experimental Psychology, supervised by Prof Kate Nation and Prof Robert Hepach. My DPhil research is on children's emotion and language development. Emotions play a pivotal role in human experiences and are closely intertwined with language structure and use in both adults and children (Lindquist, 2017). Understanding emotional language, such as the term "upset", is crucial for emotional comprehension during development (Grosse et al., 2021; Nook et al., 2020). Despite the acknowledged importance of language in emotional concept development, little research has explored the link between emotions and written language in particular. Compared to spoken language, written language is more decontextualised and linguistically more complex (Dawson et al., 2021). Written language might be instrumental in conveying nuanced emotional concepts. Hence, my DPhil research focuses on interplay between language and emotion and how this unfolds over development, with a particular focus on written language. I combined both corpus analysis to look at the emotional language in various children's language corpora; and experimental approaches to look at the influence of emotional context on language learning.

Emotion is closely associated with language, but we know very little about how children express emotion in their own writing. We used a large-scale, cross-sectional, and data-driven approach to investigate emotional expression via writing in children of different ages, and whether it varies for boys and girls. We first used a lexicon-based bag-of-words approach (after Hipson & Mohammad, 2020) to identify emotional content in a large corpus of stories (N>100,000) written by 7- to 13-year-old children. Generalized Additive Models were then used to model changes in sentiment across age and gender. Two additional approaches (BERT and TextBlob) validated and extended these analyses, converging on the finding that positive sentiments in children's writing decrease with age. These findings echo reports from previous studies showing a decrease in mood and an increased use of negative emotion words with age. We also found that stories by girls contained more positive sentiments than stories by boys. Our study shows the utility of using large-scale data-driven approaches to reveal the content and nature of children's writing. Future experimental work should build on these observations to understand the likely complex relationships between written language and emotion, and how these change over development.