<div dir="ltr"><div class="gmail_default" style="font-family:"times new roman",serif;font-size:small">Prezados</div><div class="gmail_default" style="font-family:"times new roman",serif;font-size:small"><br></div><div class="gmail_default" style="font-family:"times new roman",serif;font-size:small">Tenho uma amostra pequena com apenas 17 registros, e 7 variáveis dicotômicas, cujo resumo é:</div><div class="gmail_default"><br><pre tabindex="0" class="gmail-GGHFMYIBMOB" id="gmail-rstudio_console_output" style="outline:none;border:none;word-break:break-all;margin-top:0px;margin-bottom:0px;line-height:16px"><pre tabindex="0" class="gmail-GGHFMYIBMOB" id="gmail-rstudio_console_output" style="color:rgb(0,0,0);font-family:"Ubuntu Mono";font-size:10.4pt;white-space:pre-wrap;outline:none;border:none;word-break:break-all;margin-top:0px;margin-bottom:0px;line-height:16px"> A      B      C      D      E      F      G     
 0: 7   0:10   0: 6   0:11   0: 2   0: 1   0:15  
 1:10   1: 7   1:11   1: 6   1:15   1:16   1: 2 </pre><pre tabindex="0" class="gmail-GGHFMYIBMOB" id="gmail-rstudio_console_output" style="color:rgb(0,0,0);font-family:"Ubuntu Mono";font-size:10.4pt;white-space:pre-wrap;outline:none;border:none;word-break:break-all;margin-top:0px;margin-bottom:0px;line-height:16px"><br></pre><pre tabindex="0" class="gmail-GGHFMYIBMOB" id="gmail-rstudio_console_output" style="color:rgb(0,0,0);font-family:"Ubuntu Mono";font-size:10.4pt;white-space:pre-wrap;outline:none;border:none;word-break:break-all;margin-top:0px;margin-bottom:0px;line-height:16px">A variável G seria meu desfecho e o resto possíveis preditores</pre><pre tabindex="0" class="gmail-GGHFMYIBMOB" id="gmail-rstudio_console_output" style="color:rgb(0,0,0);font-family:"Ubuntu Mono";font-size:10.4pt;white-space:pre-wrap;outline:none;border:none;word-break:break-all;margin-top:0px;margin-bottom:0px;line-height:16px"><br></pre><pre tabindex="0" class="gmail-GGHFMYIBMOB" id="gmail-rstudio_console_output" style="color:rgb(0,0,0);font-family:"Ubuntu Mono";font-size:10.4pt;white-space:pre-wrap;outline:none;border:none;word-break:break-all;margin-top:0px;margin-bottom:0px;line-height:16px">Para veficiar a associação entre o desfecho e as outras variáveisexecutei o teste exato de fisher por ser uma amostra pequena, mas o resutados mostraram que o desfecho não depende de nenhuma variável:</pre><pre tabindex="0" class="gmail-GGHFMYIBMOB" id="gmail-rstudio_console_output" style="color:rgb(0,0,0);font-family:"Ubuntu Mono";font-size:10.4pt;white-space:pre-wrap;outline:none;border:none;word-break:break-all;margin-top:0px;margin-bottom:0px;line-height:16px"><br></pre><pre tabindex="0" class="gmail-GGHFMYIBMOB" id="gmail-rstudio_console_output" style="color:rgb(0,0,0);font-family:"Ubuntu Mono";font-size:10.4pt;white-space:pre-wrap;outline:none;border:none;word-break:break-all;margin-top:0px;margin-bottom:0px;line-height:16px"><span class="gmail-GGHFMYIBCPB gmail-ace_keyword" style="color:rgb(147,15,128);white-space:pre">> </span><span class="gmail-GGHFMYIBCOB gmail-ace_keyword" style="color:rgb(147,15,128)">fisher.test(<a href="http://data.to.work">data.to.work</a>$G, <a href="http://data.to.work">data.to.work</a>$A)
</span>
        Fisher's Exact Test for Count Data

data:  <a href="http://data.to.work">data.to.work</a>$G and <a href="http://data.to.work">data.to.work</a>$A
p-value = 0.4853
alternative hypothesis: true odds ratio is not equal to 1
95 percent confidence interval:
 0.1311443       Inf
sample estimates:
odds ratio 
       Inf 

<span class="gmail-GGHFMYIBCPB gmail-ace_keyword" style="color:rgb(147,15,128);white-space:pre">> </span><span class="gmail-GGHFMYIBCOB gmail-ace_keyword" style="color:rgb(147,15,128)">fisher.test(<a href="http://data.to.work">data.to.work</a>$G, <a href="http://data.to.work">data.to.work</a>$B)
</span>
        Fisher's Exact Test for Count Data

data:  <a href="http://data.to.work">data.to.work</a>$G and <a href="http://data.to.work">data.to.work</a>$B
p-value = 0.4853
alternative hypothesis: true odds ratio is not equal to 1
95 percent confidence interval:
 0.000000 7.625189
sample estimates:
odds ratio 
         0 

<span class="gmail-GGHFMYIBCPB gmail-ace_keyword" style="color:rgb(147,15,128);white-space:pre">> </span><span class="gmail-GGHFMYIBCOB gmail-ace_keyword" style="color:rgb(147,15,128)">fisher.test(<a href="http://data.to.work">data.to.work</a>$G, <a href="http://data.to.work">data.to.work</a>$C)
</span>
        Fisher's Exact Test for Count Data

data:  <a href="http://data.to.work">data.to.work</a>$G and <a href="http://data.to.work">data.to.work</a>$C
p-value = 0.5147
alternative hypothesis: true odds ratio is not equal to 1
95 percent confidence interval:
 0.1003871       Inf
sample estimates:
odds ratio 
       Inf 

<span class="gmail-GGHFMYIBCPB gmail-ace_keyword" style="color:rgb(147,15,128);white-space:pre">> </span><span class="gmail-GGHFMYIBCOB gmail-ace_keyword" style="color:rgb(147,15,128)">fisher.test(<a href="http://data.to.work">data.to.work</a>$G, <a href="http://data.to.work">data.to.work</a>$D)
</span>
        Fisher's Exact Test for Count Data

data:  <a href="http://data.to.work">data.to.work</a>$G and <a href="http://data.to.work">data.to.work</a>$D
p-value = 0.5147
alternative hypothesis: true odds ratio is not equal to 1
95 percent confidence interval:
 0.00000 9.96144
sample estimates:
odds ratio 
         0 

<span class="gmail-GGHFMYIBCPB gmail-ace_keyword" style="color:rgb(147,15,128);white-space:pre">> </span><span class="gmail-GGHFMYIBCOB gmail-ace_keyword" style="color:rgb(147,15,128)">fisher.test(<a href="http://data.to.work">data.to.work</a>$G, <a href="http://data.to.work">data.to.work</a>$E)
</span>
        Fisher's Exact Test for Count Data

data:  <a href="http://data.to.work">data.to.work</a>$G and <a href="http://data.to.work">data.to.work</a>$E
p-value = 1
alternative hypothesis: true odds ratio is not equal to 1
95 percent confidence interval:
 0.01971228        Inf
sample estimates:
odds ratio 
       Inf 

<span class="gmail-GGHFMYIBCPB gmail-ace_keyword" style="color:rgb(147,15,128);white-space:pre">> </span><span class="gmail-GGHFMYIBCOB gmail-ace_keyword" style="color:rgb(147,15,128)">fisher.test(<a href="http://data.to.work">data.to.work</a>$G, <a href="http://data.to.work">data.to.work</a>$F)
</span>
        Fisher's Exact Test for Count Data

data:  <a href="http://data.to.work">data.to.work</a>$G and <a href="http://data.to.work">data.to.work</a>$F
p-value = 1
alternative hypothesis: true odds ratio is not equal to 1
95 percent confidence interval:
 0.003434375         Inf
sample estimates:
odds ratio 
</pre><pre tabindex="0" class="gmail-GGHFMYIBMOB" id="gmail-rstudio_console_output" style="color:rgb(0,0,0);font-family:"Ubuntu Mono";font-size:10.4pt;white-space:pre-wrap;outline:none;border:none;word-break:break-all;margin-top:0px;margin-bottom:0px;line-height:16px"><span style="font-size:10.4pt">       Inf </span> </pre><pre tabindex="0" class="gmail-GGHFMYIBMOB" id="gmail-rstudio_console_output" style="color:rgb(0,0,0);font-family:"Ubuntu Mono";font-size:10.4pt;white-space:pre-wrap;outline:none;border:none;word-break:break-all;margin-top:0px;margin-bottom:0px;line-height:16px"><br></pre><pre tabindex="0" class="gmail-GGHFMYIBMOB" id="gmail-rstudio_console_output" style="color:rgb(0,0,0);font-family:"Ubuntu Mono";font-size:10.4pt;white-space:pre-wrap;outline:none;border:none;word-break:break-all;margin-top:0px;margin-bottom:0px;line-height:16px">Em seguida calculei o V de Cramer, Coeficiente de Contingência e phi e todos mostram independência e associação fraca.</pre><pre tabindex="0" class="gmail-GGHFMYIBMOB" id="gmail-rstudio_console_output" style="color:rgb(0,0,0);font-family:"Ubuntu Mono";font-size:10.4pt;white-space:pre-wrap;outline:none;border:none;word-break:break-all;margin-top:0px;margin-bottom:0px;line-height:16px"><br></pre><font color="#000000" face="Ubuntu Mono"><span style="font-size:13.8667px;white-space:pre-wrap">summary(assocstats(tab_cont_G_A))
# Resultado:
#                  X^2 df P(> X^2)
# Likelihood Ratio 2.3071  1  0.12878 # P-Valor > 0.05 - independentes
# Pearson          1.5867  1  0.20780 # P-Valor > 0.05 - independentes

# Phi-Coefficient   : 0.306 # Coef de Phi - Associação fraca
# Contingency Coeff.: 0.292 # Coef de Cont - Associação fraca
# Cramer's V        : 0.306 # V de Crammer - Associação fraca
</span></font><pre tabindex="0" class="gmail-GGHFMYIBMOB" id="gmail-rstudio_console_output" style="color:rgb(0,0,0);font-family:"Ubuntu Mono";font-size:10.4pt;white-space:pre-wrap;outline:none;border:none;word-break:break-all;margin-top:0px;margin-bottom:0px;line-height:16px"> </pre><pre tabindex="0" class="gmail-GGHFMYIBMOB" id="gmail-rstudio_console_output" style="color:rgb(0,0,0);font-family:"Ubuntu Mono";font-size:10.4pt;white-space:pre-wrap;outline:none;border:none;word-break:break-all;margin-top:0px;margin-bottom:0px;line-height:16px">Pelo que pesquisei este seriam os métodos mais adequados para este tipo de análise</pre><pre tabindex="0" class="gmail-GGHFMYIBMOB" id="gmail-rstudio_console_output" style="color:rgb(0,0,0);font-family:"Ubuntu Mono";font-size:10.4pt;white-space:pre-wrap;outline:none;border:none;word-break:break-all;margin-top:0px;margin-bottom:0px;line-height:16px"><br></pre><pre tabindex="0" class="gmail-GGHFMYIBMOB" id="gmail-rstudio_console_output" style="color:rgb(0,0,0);font-family:"Ubuntu Mono";font-size:10.4pt;white-space:pre-wrap;outline:none;border:none;word-break:break-all;margin-top:0px;margin-bottom:0px;line-height:16px">Há um método melhor para avaliar esse tipo de dado?</pre><pre tabindex="0" class="gmail-GGHFMYIBMOB" id="gmail-rstudio_console_output" style="color:rgb(0,0,0);font-family:"Ubuntu Mono";font-size:10.4pt;white-space:pre-wrap;outline:none;border:none;word-break:break-all;margin-top:0px;margin-bottom:0px;line-height:16px"><br></pre><pre tabindex="0" class="gmail-GGHFMYIBMOB" id="gmail-rstudio_console_output" style="color:rgb(0,0,0);font-family:"Ubuntu Mono";font-size:10.4pt;white-space:pre-wrap;outline:none;border:none;word-break:break-all;margin-top:0px;margin-bottom:0px;line-height:16px"><br></pre><pre tabindex="0" class="gmail-GGHFMYIBMOB" id="gmail-rstudio_console_output" style="color:rgb(0,0,0);font-family:"Ubuntu Mono";font-size:10.4pt;white-space:pre-wrap;outline:none;border:none;word-break:break-all;margin-top:0px;margin-bottom:0px;line-height:16px"><br></pre></pre></div>-- <br><div dir="ltr" class="gmail_signature"><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div dir="ltr"><div><font face="times new roman, serif"><i><span style="font-size:12.8px">In Jesu et Maria</span><br><br style="font-size:small"></i></font><div><div dir="ltr"><div dir="ltr"><div dir="ltr"><div style="font-size:small"><font face="times new roman, serif"><i>Obrigado</i></font></div><div style="font-size:small"><font face="times new roman, serif"><i>Prof. Elias Carvalho</i></font></div><div style="font-size:small"><font face="times new roman, serif"><i><br></i></font></div><div><font face="times new roman, serif" size="2"><i><div>"Felix, qui potuit rerum cognoscere causas" (Virgil 29 BC)</div><div>"Blessed is he who has been able to understand the cause of things"</div></i></font></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div>