@inproceedings{harrigian-dredze-2022-now,
    title = "Then and Now: Quantifying the Longitudinal Validity of Self-Disclosed Depression Diagnoses",
    author = "Harrigian, Keith  and
      Dredze, Mark",
    editor = "Zirikly, Ayah  and
      Atzil-Slonim, Dana  and
      Liakata, Maria  and
      Bedrick, Steven  and
      Desmet, Bart  and
      Ireland, Molly  and
      Lee, Andrew  and
      MacAvaney, Sean  and
      Purver, Matthew  and
      Resnik, Rebecca  and
      Yates, Andrew",
    booktitle = "Proceedings of the Eighth Workshop on Computational Linguistics and Clinical Psychology",
    month = jul,
    year = "2022",
    address = "Seattle, USA",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/add-emnlp-2024-awards/2022.clpsych-1.6/",
    doi = "10.18653/v1/2022.clpsych-1.6",
    pages = "59--75",
    abstract = "Self-disclosed mental health diagnoses, which serve as ground truth annotations of mental health status in the absence of clinical measures, underpin the conclusions behind most computational studies of mental health language from the last decade. However, psychiatric conditions are dynamic; a prior depression diagnosis may no longer be indicative of an individual`s mental health, either due to treatment or other mitigating factors. We ask: to what extent are self-disclosures of mental health diagnoses actually relevant over time? We analyze recent activity from individuals who disclosed a depression diagnosis on social media over five years ago and, in turn, acquire a new understanding of how presentations of mental health status on social media manifest longitudinally. We also provide expanded evidence for the presence of personality-related biases in datasets curated using self-disclosed diagnoses. Our findings motivate three practical recommendations for improving mental health datasets curated using self-disclosed diagnoses:1) Annotate diagnosis dates and psychiatric comorbidities2) Sample control groups using propensity score matching3) Identify and remove spurious correlations introduced by selection bias"
}