One methodological issue that I can detect in this article is the choice of forums used to draw their conclusions: > We predicted and found that anxiety, depre

One methodological issue that I can detect in this article is the choice of forums used to draw their conclusions: > We predicted and found that anxiety, depression, and suicidal ideation forums contained more absolutist words than control forums (ds > 3.14). Suicidal ideation forums also contained more absolutist words than anxiety and depression forums (ds > 1.71). I have no knowledge of this field of research, but I'm trouble understanding why its a significant result that self-oriented forums (like "anxiety, depression, and suicidal ideation forums") have more self-oriented words and absolutist words than the control forums (of which 'MumsNet' and 'StudentRoom' are mentioned in the article). In the case of 'MumsNet' at least, you'd almost certainly have a prevalence of second person pronouns.