Wojtys, Małgorzata and Marra, Giampiero and Radice, Rosalba (2018) Copula based generalized additive models for location, scale and shape with non-random sample selection. Computational Statistics and Data AnalysiS 127 , pp. 1-14. ISSN 0167-9473.
|
Text
GSSM paper.pdf - Author's Accepted Manuscript Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (444kB) | Preview |
Abstract
Non-random sample selection is a commonplace amongst many e mpirical studies and it appears when an output variable of interest is available only for a restricted non- random sub-sample of data. An extension of the generalized additive models for location, scale and shape which account s for non-random sample selection by introducing a selection equation is discussed. The proposed approach all ows for potentially any parametric distribution for the outcome variable, any parametric link function for the sele ction equation, several dependence structures between the (outcome and selection) equations through the use of copula e, and various types of covariate effects. Using a special case of the proposed model, it is shown how the score equation s are corrected for the bias deriving from non-random sample selection. Parameter estimation is carried out with in a penalized likelihood based framework. The empirical effectiveness of the approach is demonstrated through a sim ulation study and a case study. The models can be easily employed via the gjrm() function in the R package GJRM .
Metadata
| Item Type: | Article |
|---|---|
| Keyword(s) / Subject(s): | additive predictor, copula, marginal distribution, non-r andom sample selection, penalized regression spline, simultaneous equation estimation |
| School: | Birkbeck Faculties and Schools > Faculty of Business and Law > Birkbeck Business School |
| Depositing User: | Rosalba Radice |
| Date Deposited: | 27 Apr 2018 09:00 |
| Last Modified: | 02 Aug 2023 17:41 |
| URI: | https://eprints.bbk.ac.uk/id/eprint/22236 |
Statistics
Additional statistics are available via IRStats2.
Archive Staff Only (login required)
