It provides tools for mapping proportions and indicators defined on the unit interval, widely used to measure, for instance, unemployment, educational attainment and also disease prevalence. It implements Beta-based small area methods, particularly indicated for unit interval responses, comprising the classical Beta regression models, the Flexible Beta model and Zero and/or One Inflated extensions. Such methods, developed within a Bayesian framework, come equipped with a set of diagnostics and complementary tools, visualizing and exporting functions. A customized parallel computing is built-in to reduce the computational time. The features of the tipsae package assist the user in carrying out a complete SAE analysis through the entire process of estimation, validation and results presentation, making the application of Bayesian algorithms and complex SAE methods straightforward. A Shiny application with a user-friendly interface can be launched to further simplify the process.
Silvia De Nicolò, silvia.denicolo@unibo.it
Aldo Gardini, aldo.gardini@unibo.it
JSStipsae
rstantipsae
carpenter2017stantipsae
janicki2020propertiestipsae
vehtari2017practicaltipsae
datta2011bayesiantipsae
kish1992weightingtipsae
fabrizi2011hierarchicaltipsae
morris2019bayesiantipsae
DeNicolo2021tipsae
shinytipsae