Re: [ABE-L]: Sim, uma conciliação é possível

[Hedibert Freitas Lopes <hlopes@chicagogsb.edu>]

Queria humildemente fazer uma proposta:  criar uma lista ABE-piadas (e' possivel Lucia?) onde a gente podia de vez em quando visitar para rir um pouco, ser feliz, fazer chacota um dos outros, enfim tudo em nome da paz e da preservacao da ABE-L.  Por favor, isso nao e' uma piada.

Atenciosamente,
Hedibert


----- Original Message -----
From: Francisco Cribari <cribari@de.ufpe.br>
Date: Wednesday, September 19, 2007 3:56 pm
Subject: [ABE-L]: Sim, uma conciliação é possível
To: abe-l@ime.usp.br


> [*** O Ministério da Saúde adverte: quem não tiver senso de humor, por
>  favor vá direto para a tecla "Delete". ***] 
>  
>  
>  Inspirado pelo paper do Fraser, que aborda três problemas importantes 
> a
>  partir de metodologias distintas, decidi compilar uma lista de artigos
>  (ver abaixo) que abordam um problema atual e de elevado interesse no
>  nosso país. Notem que os artigos usam metodologias distintas, mas, em
>  sua grande maioria, chegam à mesma inferência, o que sugere que
>  correntes diferentes do pensamento estatístico não precisam ser
>  necessariamente vistas como excludentes. 
>  
>  A lista dos artigos: 
>  
>  [1] Pacheco, Cecílio. (2007). Who killed Taís Grimaldi? A Bayesian
>  proof. Submitted for publication. 
>  
>  Abstract: In this paper we reveal who killed Taís Grimaldi. The proof 
> is
>  divided into three separate sections. Using Bayesian arguments, we first
>  show that Taís Grimaldi is indeed dead. In the second segment of the
>  proof, we prove that God exists. Finally, the proof is completed by a
>  conversation with God, in which He reveals who killed Taís Grimaldi. 
> In
>  future research we shall talk to Him about other stuff. 
>  
>  [2] Caramba, Felício. (2007). Who killed Taís Grimaldi? An answer from
>  higher order asymptotics. Submitted for publication.
>  
>  Abstract. We solve the problem by letting the number of observations 
> go
>  to infinity, i.e., we kill all of the remaining soap opera characters.
>  After that step, they are all dead and there is no need to know who
>  killed Taís Grimaldi, which completes the proof. 
>  
>  [3] Camarada, Fernando. (2007). Who killed Taís Grimaldi? An answer from
>  empirical Bayes. Submitted for publication.
>  
>  Abstract: According to both die-hard Bayesians and frequentists (see
>  Pacheco, 2007, and Caramba, 2007), Taís Grimaldi is dead. Therefore, 
> she
>  is dead. 
>  
>  [4] Sampaio, Dagoberto. (2007). Is Taís Grimaldi really dead? A
>  counter-proof from objective Bayes. Submitted for publication.
>  
>  Abstract: The proof can be summarized as follows. In the soap opera,
>  Taís Grimaldi is portrayed by Alessandra Negrini, who is alive.
>  Therefore, Taís Grimaldi is alive. And even better: this proof pisses
>  off both Bayesians and frequentists. 
>  
>  [5] Cabrini, Fernando. (2007). Who killed Taís Grimaldi? An answer using
>  bootstrap. Submitted for publication. 
>  
>  Abstract: We resample Taís Grimaldi B times (B is a large positive
>  integer) and we kill her on all B occasions. We are, thus, 100% sure
>  that she is dead. And we know who killed her: it was the the damn
>  bootstrap!
>  
>  --
>  Francisco Cribari-Neto               voice: +55-81-21267425
>  Departamento de Estatistica          fax:   +55-81-21268422
>  Universidade Federal de Pernambuco   e-mail: cribari@de.ufpe.br
>  Recife/PE, 50740-540, Brazil         web: www.de.ufpe.br/~cribari
>  
>    "I took a speed reading course and read 'War and Peace' in twenty
>     minutes. It involves Russia."                      --Woody Allen
>  
>  
>