Those single scores we use to rank people on things like intelligence might actually be mathematical illusions.
March 24, 2026
Original Paper
Refactor Analysis: Predictive Evaluations of Factor Models and Dimensionality
arXiv · 2603.20938
The Takeaway
Researchers found that the standard tests used to justify 'one-number' summaries (like IQ or performance rankings) often appear to fit data perfectly while failing to actually predict real-world responses. This suggests that many of our most trusted scientific metrics might be fundamentally flawed.
From the abstract
Unidimensional factor models justify some of the most consequential summaries in science -- single scores, single ranks, and single leaderboards -- yet unidimensionality is usually assessed indirectly by fitting and evaluating models on images of the data (e.g., correlation matrices) rather than on the response matrix itself. We introduce Refactor analysis, a data-first evaluation paradigm that converts a one-factor solution into a rank-1 prediction of the original matrix by estimating both resp