Using text to predict psychological and physical health: A comparison of human raters and computerized text analysis
作者：
Highlights：
• Compares two methods of text analysis in predicting self-reported health.
• Human ratings of text were better at predicting depression.
• Automated text analysis was sufficient for predicting pain catastrophizing.
• Automated text analysis was sufficient for predicting illness intrusiveness.
• Neither method was sufficient for predicting life satisfaction or pain severity.
摘要
•Compares two methods of text analysis in predicting self-reported health.•Human ratings of text were better at predicting depression.•Automated text analysis was sufficient for predicting pain catastrophizing.•Automated text analysis was sufficient for predicting illness intrusiveness.•Neither method was sufficient for predicting life satisfaction or pain severity.
论文关键词：Chronic pain,Expressive writing,Text analysis,Sentiment analysis,Human coding
论文评审过程：Received 29 September 2016, Revised 9 May 2017, Accepted 28 June 2017, Available online 30 June 2017, Version of Record 15 July 2017.
论文官网地址：https://doi.org/10.1016/j.chb.2017.06.038
