Smooth varying-coefficient models in Stata

• Autores: Fernando Rios-Avila
• Localización: The Stata journal, ISSN 1536-867X, Vol. 20, Nº. 3, 2020, págs. 647-679
• Idioma: inglés
• Texto completo no disponible (Saber más ...)
• Resumen

  • Nonparametric regressions are powerful statistical tools that can beused to model relationships between dependent and independent variables withminimal assumptions on the underlying functional forms. Despite their poten-tial benefits, these models have two weaknesses: The added flexibility creates acurse of dimensionality, and procedures available for model selection, like cross-validation, have a high computational cost in samples with even moderate sizes.An alternative to fully nonparametric models is semiparametric models that com-bine the flexibility of nonparametric regressions with the structure of standardmodels. In this article, I describe the estimation of a particular type of semipara-metric model known as the smooth varying-coefficient model (Hastie and Tibshi-rani, 1993,Journal of the Royal Statistical Society, Series B55: 757–796), basedon kernel regression methods, using a new set of commands withinvcpack. Thesecommands aim to facilitate bandwidth selection and model estimation as well ascreate visualizations of the results.