Reading focus
Syllabus: How Should We Classify Psychopathology?
Compare what the diagnostic and dimensional predictor sets contain and what each adds.
Why you’re reading it. Evaluate a substantive empirical test after the implementation proposal.
A question to consider. Does better prediction of one twenty-year outcome establish that clinicians should replace DSM diagnoses with HiTOP? Specify what this study does and does not test.
Why this reading
Competing classifications need comparisons on something beyond their own structure. This longitudinal study asks whether dimensions associated with HiTOP and DSM diagnostic groups provide different information about later outcomes.
Before you read
Incremental prediction asks whether one set of measurements adds information after accounting for another. Explained variance is a sample-based performance measure. A comparison in the same dataset is different from testing a forecasting rule in independent data.
Read for
- Track the six-month predictors, twenty-year outcomes and losses to follow-up.
- Notice that the dimensional comparison includes functional impairment, and that combining approaches can retain information from both.
Consider afterwards
- What would make a comparison unfairly favor one approach?
- Why would improved prediction still leave a causal explanation and a benefit from clinical adoption unproven?
Editorial context
The analysis concerns 316 participants from one first-admission psychosis cohort. Four symptom dimensions plus impairment do not represent every HiTOP domain. These regression comparisons are not held-out clinical forecasts; outcome-specific results and retained DSM information matter. The study does not test whether clinicians using HiTOP improve patient outcomes.