Regression Models: Censored, Sample Selected, Or Truncated Data
"What techniques can social scientists use when an outcome variable for a sample (for example, y) is not representative of the population for which generalized results are preferred? Author Richard Breen provides an introduction to regression models for such data, including censored, sample-selected, and truncated data. Regression Models begins with a discussion of the Tobit model and examines issues such as maximum likelihood estimation and the interpretation of parameters. The author next discusses the basic sample selection model and the truncated regression model. Elaborating on the modeling of censored and sample-selected data via maximum likelihood, he shows the close links between the models introduced and other regression models for non-continuous dependent variables, such as the ordered probit. Concluding with an exploration of some of the criticisms of these approaches and difficulties associated with them, this volume gives readers a guide to the practical utility of these models."--Publisher description.
What people are saying - Write a review
We haven't found any reviews in the usual places.
The Tobit Model for Censored Data
SampleSelection Models and
Extensions of the Basic Models
The Expected Value of
About the Author
Analysis assume assumptions asymptotically bivariate normality censored and sample-selected censored data censored regression model censored sample censoring threshold Chapter Chesher and Irish compute conditional expectation correct correlation covariance denote density function dependent variable distributed random variable effect equal error term evaluated example exchange rate expected value explanatory variables expression financial assets given Hagan Heckman two-step method heteroscedasticity homoscedasticity households income inverse Mill's ratio latent variable likelihood function linear log-likelihood function Maddala matrix ML estimates models for censored nonnormal nonzero normal distribution function normally distributed random OLS regression ordered probit outcome equation parameter estimates partial derivative population problem random variable sample-selected data sample-selection bias sample-selection model selection and outcome selection bias selection equation selection model shown in Equation simply social science spend standard errors subsample test for normality tion Tobit coefficients Tobit model total wealth truncated data truncated regression model two-stage two-step approach two-step estimator unbiased uncensored observations vector