The Gottschalk-Gleser Depression Scale quantifies depressive thematic content in speech, clause by clause, across seven subscales. It is a well-validated 50-year-old content-analysis instrument (Gottschalk & Hoigaard-Martin 1986, extending the Gottschalk-Gleser content-analysis lineage) — purpose-built for transcript annotation: each scorable clause is a span an annotator (or the agent) tags with a subscale.
A distinctive feature is perspective-based weighting: content the
speaker attributes to themselves (weight 3) counts more heavily than
content attributed to others (2) or to impersonal situations / denial
(1); active suicidal content (in Self-accusation) reaches weight 4.
Hopelessness, Psychomotor retardation, and Somatic concerns use a flat
weight of 1. Per-subscale scores are computed as
sqrt((weighted_sum + 0.5) x 100 / word_count) and summed for a total
depression score.
These curated skills are the first modern LLM operationalization of the instrument, developed on the DAIC-WOZ clinical-interview corpus.
Attribution. The Gottschalk-Gleser method is summarized here from published sources; the original content-analysis manual is © 1969 The Regents of the University of California (University of California Press). Tagaroo's curated scoring skills are an independent operationalization and reproduce no proprietary manual content.