EGA-DMD estimates item parameters thousands of times faster than MIRT, new simulation shows
GolinoHudson · x · 2026-08-15
A new simulation study compares Exploratory Graph Analysis (EGA-DMD) with Multidimensional Item Response Theory (MIRT) for item parameter estimation. Results show EGA-DMD is orders of magnitude faster (median runtime 0.12s vs 702s at N=5000), with near-perfect convergence and comparable recovery of difficulty and discrimination. Under strong factor correlations and cross-loadings, EGA-DMD converges to its own estimand, indicating different representations of multidimensional item geometry.
More from Research
- Discussion on Definitions of Mechanistic vs Causal Models — sirbayes · 2026-08-15
- New England Mech Interpretability workshop kicks off — benno_krojer · 2026-08-15
- Paper Advocates Aligning Language Models with Long-Term User Well-being — 996roma · 2026-08-15
- Paper reveals LM Head as gradient bottleneck amid Anthropic vocab size rumors — JFPuget · 2026-08-15
- Weights Unknown Conference Focuses on Scaling and Agents — DynamicWebPaige · 2026-08-15
- Epoch researcher lists top questions governing AI's global impact — Jsevillamol · 2026-08-15