Limited dependent poisson regression

• The paper introduces aspects of statistical inference in Poisson regression models, with a dependent variable subject to truncation and/or censoring. Losses in efficiency, due to censoring and truncation, of the maximum likelihood estimator are illustrated. Predictors and predictor variances are given. New adaptations to tests for unobserved random heterogeneity (overdispersion) are given. Power properties are examined in a small Monte Carlo experiment. A labour market illustration is given.

Ämnesord

SAMHÄLLSVETENSKAP  -- Ekonomi och näringsliv -- Nationalekonomi (hsv//swe)

SOCIAL SCIENCES  -- Economics and Business -- Economics (hsv//eng)

Nyckelord

ekonometri

Econometrics

Publikations- och innehållstyp

ref (ämneskategori)

art (ämneskategori)