作者ppta (幸福的瞬间)
看板Statistics
标题[问题] ODDS RATIO及卡方的问题.......
时间Fri May 18 22:05:44 2007
http://myurl.com.tw/ogck 是我要问的问题TABLE3表格，如果太小，烦请各位高手们
下载後放大，感恩。
我看我的PAPER，上面写着：
As shown in Table 3, pooling the results from these eight studies yielded the
results
(OR=1.27, z=1.88, p =0 .06), which suggests a trend or
significant evidence that the s allele of 5-HTTLPR is associated
with suicide attempts in patients with psychiatric diseases. In the
analysis of the same group of subjects (Table 3), there is a
significant association between the genotypes and suicide attempt
(OR=1.55, z= 2.86, p= .004)。
结论就是都有显着差异，我好奇说，明明是要算POOLED OR，怎麽又会扯到P值，请问
这个P值怎麽算出来的，我有SPSS这套软体，能教教我吗？
但是我又有下载一个软体
http://www.medcalc.be/manual/meta-oddsratio.php
也是能算出POOLED OR，但是上述网页提及
The Total Odds ratio with 95% CI is given both for the Fixed effects model
and the Random effects model. If the value 1 is not within the 95% CI, then
the Odds ratio is statistically significant at the 5% level (P<0.05).
也就是若95% CI的OR值若不包含1，才算有显着标准，才代表观察操作组与控制组有显
着差异，是吧？以这个标准来看，TABLE3的ALLELE部分（左边）有包含1在内，应该是没
有显着差异，为什麽文章却说有呢？到底那一个是对的？
如果有高手了解为什麽，可教我用SPSS算P值吗？
谢谢
--
※ 发信站: 批踢踢实业坊(ptt.cc)
◆ From: 203.71.2.194