Informed Consent

About this study

LongCounsel-8 is a dataset of longitudinal synthetic counseling conversations with session-level PHQ-8 depression labels. This study evaluates whether its language resembles real counseling and whether PHQ-8 symptom labels are reflected in the conversation text in a way comparable with real data.

Privacy and use of responses

We do not ask for your name, contact details, or other direct identifiers. A random browser ID stores your progress. The study evaluates the dataset, not you: your answers will not be used to assess your personality, mental health, or decision-making.

Responses, response times, task completion, and—when applicable—the self-reported general or expert participant group are de-identified and used to evaluate the dataset's realism and validity. Reimbursement is handled separately using a random completion code.

Content warning

Some passages describe depression, severe emotional distress, or self-harm. If this content may trigger or distress you, please do not participate. You may pause or leave the study at any time.