Why this reading
The taxometric literature is the most direct attempt to test whether a psychological construct is categorical or continuous at the latent level. This review gathers that literature, weighs it by methodological quality, and reports what the accumulated results favour. The answer is largely dimensional, with a small number of constructs where evidence for a genuine category remains. Reading the review is the quickest way to understand both what taxometric methods can show and why their earlier, more categorical findings should be read with care.
Before you read
Taxometric methods, developed by Paul Meehl, ask whether the pattern of relationships among indicators is more consistent with a latent category (a taxon) plus a complement class, or with a continuum. The review explains the procedures briefly; it is enough to grasp the question and that the methods depend on assumptions about the indicators.
Notice the difference between the structure of a construct in the population and the way clinicians use thresholds. The review is about the former.
Read for
- How the authors decide which studies to include and how they rate methodological quality. Which features make a taxometric study credible?
- The pattern across time: what happened to the proportion of categorical findings as methods improved, and how the authors interpret that.
- The constructs for which categorical findings persist, and what the authors say about them.
- What the review implies for diagnostic manuals that treat every disorder as a category.
Consider afterwards
- What would a genuine category look like in data, and what could produce a spurious one?
- If most constructs are dimensional, does that settle how clinicians should decide who receives treatment? What else would you need to know?
Editorial context
This review is retained as an optional predecessor to Haslam, McGrath, Viechtbauer, and Kuppens (2020). The later meta-analysis extends the evidence and gives particular attention to simulation-based comparisons. Neither review determines clinical thresholds from latent structure alone. Read the older review to follow the development of the argument; use the later synthesis for the current core reading.