Erling B. Andersen

Overview

Erling B. Andersen is a statistician at the Department of Statistics, University of Copenhagen. His research focuses on latent variable models, item response theory, and discrete statistical methods with applications in psychometrics and the social sciences. He is known for foundational work on maximum likelihood estimation in the Rasch model and for synthesizing the latent structure analysis literature.

Key Contributions

• Andersen (1982): "Latent Structure Analysis: A Survey" (Scandinavian Journal of Statistics 9(1): 1–12); invited survey covering the field from Lazarsfeld (1950) through the EM algorithm (Goodman 1974/DLR 1977); covers latent class models, item characteristic curves, latent profile analysis, multi-population comparison, and latent factor analysis. See Andersen (1982).
• Andersen and Madsen (1977): "Estimating the Parameters of the Latent Population Distribution" (Psychometrika 42: 357–374) — Newton-Raphson ML estimation of $(\mu_\theta,\sigma_\theta^2)$ in the pure Rasch model using marginal likelihood; established numerical feasibility of continuous latent variable IRT.
• Andersen (1980a): Discrete Statistical Models with Social Science Applications (North-Holland) — monograph covering the Rasch model, latent class models, and related discrete-data methods.
• Andersen (1980b): "Comparing Latent Distributions" (Psychometrika 45: 121–134) — multi-population model comparing $N(\mu_l,\sigma_l^2)$ latent distributions; tests $\mu_1=\cdots=\mu_L$ for whether populations share a common latent level.

Related

• Latent Class Model
• EM Algorithm
• Rasch Model
• Andersen (1982) Latent Structure Analysis: A Survey