Limited Dependent Variable Modelling (Truncated and censored Regression models) with Application

Nawzad. M. Ahmed

doi:10.25098/2.2.32

Authors

Nawzad. M. Ahmed

DOI:

https://doi.org/10.25098/2.2.32

Keywords:

Censored Regression Model (CRM), Maximum Likelihood Estimator (MLE), Marginal Effect (ME), Ordinary Least Square (OLS)

Abstract

this study is about fitting a such regression model for Tobit (Truncated), and (censored) data regression models for a sample of persons (n=500), with their sugar rate per person as a response variable (Y), and explanatory variables (X₁: Age , X₂: Rate of cholesterol.gm/dl, and X₃: Triglyceride.gm/dl). (Data of this study were obtained through the follow-up of diabetic patients directly from the laboratories of analysis of blood diseases and diabetes in several laboratories with experience in the city of Sulaymaniyah). The persons whom their sugar rate is above 120gm/dl only remained and the others are censored as a zero values (left censored), also the same set of data was truncated from below, then in truncated sample, only the cases under risk of diabetes (greater than the sugar rate=120gm/dl) are recorded, and the others are omitted from the data set. With remaining explanatory with their actual records, and then after fitting censored and truncated models also in order to detect the unbiased, and consistency of the estimated (Parameters) model, the (marginal effects for explanatory) are calculated and comparing these two models with multiple regression model estimation by an ordinary least square method (OLS).

References

- Alferd. D, and Bowling, G. “Regression with Social Data Modelling Continues and Limited Response Variables”,(2004). State University, Dept. of Sociology, Ohio, John Wiley & Sons, Inc., Publication, Hoboken, New Jersy.

- Franses, P. H and Richard. P,(2001). “Quantitative models in Marketing Research “Cambridge University Press.

- Green, W. H, “Econometric Analysis” (2003). Fifth Edition, Upper Saddle River, NJ, Prentice Hall.

- Long, J. S. “Regression Models for Categorical and Limited Dependent Variables” (1997). Thousand Oaks, CA: Saga Publication.

- Maddala, G. S. “Limited Dependent and Qualitative Variables in Econometrics” (1983). Cambridge University Press, Cambridge, UK. ISBN 0-521-33825-5. OCLC 25207809.

- Richard Breen. “Regression Models Censored Sample Selected or Truncated Data”, (1996). Saga Publication Ltd. California 9130, Email:order@sagpub.com, ISBN 0-8039-5710-6.

- Stock, James H.; Watson, Mark W. “Introduction to Econometrics” (2003). Addison-Wesley, Bosten. ISBN 0-201-71595-3.

- Tobin, J. “Estimation of relationships for limited dependent variables” (1958), Econometrica. 26(1): P:(24-36).

- Toomet. O, Henningsen. A” Tools for Maximum Likelihood Estimation”. (2010). R package version 0.7, http://CRAN.R-project.org/package=maxLik. maxLik.

- Wooldridge, J. M. “Econometric Analysis of Cross Section and Panel Data” (2002). MIT Press, Cambridge. ISBN 0-262-23219-7. OCLC 47521388.