<p>This paper describes how sentiment analysis was used to investigate changes in adolescents’ feelings during the pandemic, as well as their longitudinal relationship with internalizing problems. Participants included 750 students in Southern British Columbia, Canada (<i>M</i><sub><i>age</i></sub>=14.66 years, <i>SD</i> = 1.17 years; 46.8% girls). Adolescents self-report their internalizing problems (e.g., anxiety, depression, and stress) and also included an open-ended question that asked participants to describe how they feel personally since the start of COVID-19 at both Wave 1 (October and December 2020) and Wave 2 (October and December 2021). We calculated the National Research Council Canada (NRC) Emotion value, which categorized the responses according to eight basic emotions (e.g., anger, joy, fear) at both time points. Cross-lagged analysis was then used to investigate the longitudinal relations between negative emotion expression and internalizing problems. Results showed that adolescents had higher levels of trust and anticipation but also demonstrated some fear, sadness, and anger at both time points. As compared to W1, adolescents showed lower levels of negative emotion and higher levels of positive emotion at W2. Our results from the cross-lagged model indicated that adolescents’ internalizing problems contributed to later increases in negative emotion expression.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Canadian adolescents’ feelings and internalizing problems during the pandemic: a longitudinal sentiment analysis

  • Bowen Xiao,
  • Jennifer D. Shapka

摘要

This paper describes how sentiment analysis was used to investigate changes in adolescents’ feelings during the pandemic, as well as their longitudinal relationship with internalizing problems. Participants included 750 students in Southern British Columbia, Canada (Mage=14.66 years, SD = 1.17 years; 46.8% girls). Adolescents self-report their internalizing problems (e.g., anxiety, depression, and stress) and also included an open-ended question that asked participants to describe how they feel personally since the start of COVID-19 at both Wave 1 (October and December 2020) and Wave 2 (October and December 2021). We calculated the National Research Council Canada (NRC) Emotion value, which categorized the responses according to eight basic emotions (e.g., anger, joy, fear) at both time points. Cross-lagged analysis was then used to investigate the longitudinal relations between negative emotion expression and internalizing problems. Results showed that adolescents had higher levels of trust and anticipation but also demonstrated some fear, sadness, and anger at both time points. As compared to W1, adolescents showed lower levels of negative emotion and higher levels of positive emotion at W2. Our results from the cross-lagged model indicated that adolescents’ internalizing problems contributed to later increases in negative emotion expression.