Machine learning of language use on Twitter reveals weak and non-specific predictions
Bibliographic record
Abstract
Abstract Depressed individuals use language differently than healthy controls and it has been proposed that social media posts can be used to identify depression. Much of the evidence behind this claim relies on indirect measures of mental health and few studies have tested if these language features are specific to depression versus other aspects of mental health. We analysed the Tweets of 1006 participants who completed questionnaires assessing symptoms of depression and 8 other mental health conditions. Daily Tweets were subjected to textual analysis and the resulting linguistic features were used to train an Elastic Net model on depression severity, using nested cross-validation. We then tested performance in a held-out test set (30%), comparing predictions of depression versus 8 other aspects of mental health. The depression trained model had modest out-of-sample predictive performance, explaining 2.5% of variance in depression symptoms ( R 2 = 0.025, r = 0.16). The performance of this model was as-good or superior when used to identify other aspects of mental health: schizotypy, social anxiety, eating disorders, generalised anxiety, above chance for obsessive-compulsive disorder, apathy, but not significant for alcohol abuse or impulsivity. Machine learning analysis of social media data, when trained on well-validated clinical instruments, could not make meaningful individualised predictions regarding users’ mental health. Furthermore, language use associated with depression was non-specific, having similar performance in predicting other mental health problems.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".