John Philipe Hinde | The National University Ireland/Cork - Irlanda

• Processo: 03/02244-2
• Modalidade de apoio: Auxílio à Pesquisa - Pesquisador Visitante - Internacional
• Vigência: 07 de junho de 2003 - 28 de junho de 2003
• Área do conhecimento: Ciências Exatas e da Terra - Probabilidade e Estatística - Probabilidade e Estatística Aplicadas
• Pesquisador responsável: Clarice Garcia Borges Demétrio
• Beneficiário: Clarice Garcia Borges Demétrio
• Pesquisador visitante: John Philipe Hinde
• Inst. do pesquisador visitante: National University of Ireland, Dublin (NUI), Irlanda
• Instituição Sede: Escola Superior de Agricultura Luiz de Queiroz (ESALQ). Universidade de São Paulo (USP). Piracicaba , SP, Brasil
• Assunto(s): Superdispersão Intercâmbio de pesquisadores Colaboração científica
• Palavra(s)-Chave do Pesquisador: Discrete Data | Multinomial Data | Overdispersion | Zero Inflation

Resumo

The aim of Johns visit is to build on previous collaborative work between us. There are a number of sírands of ongoing work that we are kecn to bring to completion, while at the same time starting out on new avenues of associated work. The most important aspect is to make substantial progress on our monograph Overdispersion: Models and Estimation - a further period of intense collaborative work is nccessary to move this project forward. At the same time as working on this, there will be the opportunity to review exisiting collaborative research work on overdispersion and zero-inflation, the later work being jointly with Martin Ridout, University of Kent, UK. As well as attempting to tie up a number of previous areas of work, we will identiíy additional aspects for further collaborative research possibly Ieading to a student visiting NUI, Galway under a postgraduate sandwich programme scheme with support from CAPES. Indeed, it is plarmed that Dr. de Freitas (a previous sandwich student jointly supervised by us) will also visit ESALQ/USP for the final week of John's visit, to work on the joint paper Modelling Overdispersion in Multinomial Data, and to consider new avenues for joint work. One possibility is to consider models for multinomial data with many zeros - data of this type is very common in many species abundance studies on ecology and related areas, however, at present there is no satisfactory method of handling such data in statistical modelling. The visit will be complemented by Professor Demétrio's trip to Europe later in the summer, which will include a visit to Martin Ridout. All these activities are part of the Research Project "Superdispersão em dados discretos" already approved by FAPESP (Processo n° 01/10035-9). (AU)