ABSTRACT
In recent decades, there has been an increasing interest in the relation between lexical features and texts of psychological states. Previous studies demonstrated that some lexical features varied significantly among the texts of psychological states. However, the lexical features at the textual level have received little attention. This paper extends this work by examining the performance of quantitative linguistic indices in classifying texts of psychological issues. A large dataset of forum posts including texts of anxiety, depression, suicide ideation, and normal states were experimented with Machine Learning algorithms. The results revealed that the quantitative linguistic indices with Machine Learning algorithms achieved a high level of success in identifying psychological states. Meanwhile, some quantitative linguistic indices, namely, ALT and Writer’s view, may extract adequate lexical features for classifying texts of different psychological states. The study is probably the first attempt that uses quantitative linguistic indices as lexical features to detect texts of psychological states, and the findings may contribute to our understanding of how accuracy may be enhanced in the identification of various psychological states. Finally, the implications of these findings are discussed.
Publisher’s Note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Acknowledgments
We thank the JQL referees and the editors for their insightful comments. Their suggestions have significantly enhanced the quality of the initial manuscripts.
Disclosure Statement
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Data Availability Statement
Publicly available datasets were analysed in this study. This data can be found here: We used AlMosaiwi and Johnstone’s (2018) dataset which can be accessed at https://doi.org/10.6084/m9.figshare.474 3547.v1.
Supplemental data
Supplemental data for this article can be accessed online at https://doi.org/10.1080/09296174.2023.2256211.
Notes
1. The dataset can be accessed at https://doi.org/10.6084/m9.figshare.4743547.