[R-br] GLM

ana paula coelho madeira apcmadeira em hotmail.com
Quinta Março 19 15:15:20 BRT 2015


Boa tarde!

Tenho um experimento que visa analisar como transgenia e controle químico (tratamentos)  afetam número
de parasitóides (Chelonus).
Construí o modelo:

m1<-glm.nb(n_parasit_Chelonus~trat)  
> summary(m1) 

Coefficients:
              Estimate Std. Error z value Pr(>|z|)    
(Intercept)    2.54553    0.16046  15.864  < 2e-16 ***
tratC         -0.24295    0.23851  -1.019   0.3084    
tratD          0.17905    0.21973   0.815   0.4152    
tratF        -21.84812 4713.30843  -0.005   0.9963    
tratG          0.46262    0.21032   2.200   0.0278 *  
tratI         -1.22378    0.31394  -3.898 9.69e-05 ***
tratJ          0.12862    0.22165   0.580   0.5617    
tratL        -21.84812 4713.30843  -0.005   0.9963    
tratM         -0.40547    0.24759  -1.638   0.1015    
tratN          0.28768    0.21586   1.333   0.1826    
tratP         -3.23868    0.72931  -4.441 8.96e-06 ***
tratQ         -0.04001    0.22868  -0.175   0.8611    
tratS        -21.84812 4713.30843  -0.005   0.9963    
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for Negative Binomial(40.7143) family taken to be 1)

    Null deviance: 376.329  on 51  degrees of freedom
Residual deviance:  39.605  on 39  degrees of freedom
AIC: 234.09


Como posso interpretar esses resultados?

Agradeço pela ajuda.

Ana
 		 	   		  
-------------- Próxima Parte ----------
Um anexo em HTML foi limpo...
URL: <http://listas.inf.ufpr.br/pipermail/r-br/attachments/20150319/4c295bf2/attachment.html>


Mais detalhes sobre a lista de discussão R-br