w*****1
发帖数: 473 | 1 【 以下文字转载自 Statistics 讨论区 】
发信人: wz99331 (dotti), 信区: Statistics
标 题: How to use sas to calculate the power of logistic regressi
发信站: BBS 未名空间站 (Tue Nov 4 11:28:25 2014, 美东)
To generate the statement like "given the sample size we have a %90 chance
of detecting an x percent or some unit difference of incidence of heart
defect between high fat and normal diet group at a p value of 0.05. the
sample size for high fat is 516, for normal diet is 2262. Could anyone help
me? Thanks a lot! |
|
w*****1
发帖数: 473 | 2 【 以下文字转载自 Statistics 讨论区 】
发信人: wz99331 (dotti), 信区: Statistics
标 题: How to use sas to calculate the power of logistic regressi
发信站: BBS 未名空间站 (Tue Nov 4 11:28:25 2014, 美东)
To generate the statement like "given the sample size we have a %90 chance
of detecting an x percent or some unit difference of incidence of heart
defect between high fat and normal diet group at a p value of 0.05. the
sample size for high fat is 516, for normal diet is 2262. Could anyone help
me? Thanks a lot! |
|
|
|
s*****n
发帖数: 3416 | 5 一个猎头发给我的。
从entry 到senior都可以，如果有兴趣，站内联系我，我给你猎头的联系方法。
我不负任何责任啊。
Experience and Skills
• Minimum of a Bachelors degree. While we don’t require a specific
degree, typical successful candidates have a degree in Statistics, Computer
Science, Engineering, Psychology, Marketing, or Geography.
• A Masters degree is preferred, but not required.
• Knowledge of analytical techniques. Specifically working knowledge
of concepts such as, clustering, segmentation and time-series (i.e.
regressi |
|
t********y
发帖数: 469 | 6 one can check for multicollinearity by means of the correlation matrix. In
such a matrix when the correlation coefficient between two explanatory
variables is above 0.8 one needs to be aware of possible collinearity. If
the correlation coefficient is above 0.95 the problem is really serious.
A diagnostic approach to check for multicollinearity after performing
regression analysis is to display the Variance Inflation factor (VIF).
VIF is a measure of how much the variance of an estimated regressi |
|
|
j*******2
发帖数: 309 | 8 //fit your model first
logit y x1 x2, [options]
//save your model
snapshot save, label("prediction model")
//then set the predictors the value you want: normally categorical variable
to the reference level, continuous variable to its mean
//predict
predict prob |
|
a*********n
发帖数: 1331 | 9 super many thanks! I am going to run it now... |
|
a*********n
发帖数: 1331 | 10 如果是好几个independent variables,也是用这些code吗？
多谢了! |
|
a*********n
发帖数: 1331 | 11 " //then set the predictors the value you want: normally categorical
variable
这步怎么做？
sorry,我不熟悉stata,现在处在急来抱佛脚阶段。非常感谢！
variable |
|
|
j*******2
发帖数: 309 | 13 logit *********
snapshot********
collapse (mean) x1
gen x2=*
//restore your model back
snapshot restore
baozi, plz~~~~~~~ |
|
|
|
j*******2
发帖数: 309 | 16 yes, looks like it's easier than my code.I should have used this one. |
|
a*********n
发帖数: 1331 | 17 我用margins怎么总是得到error message:
"
factor ethnic not found in e(b)
"
问题在哪？
我的margins命令 “margins ethnic, atmeans " |
|
a*********n
发帖数: 1331 | 18 你这个出来的结果好像不太对，结果是每个observation的probability，但是我要的是
每个independent variable的每个category的probability. 怎么改才对？
variable |
|
|
|
w*****1
发帖数: 473 | 21 To generate the statement like "given the sample size we have a %90 chance
of detecting an x percent or some unit difference of incidence of heart
defect between high fat and normal diet group at a p value of 0.05. the
sample size for high fat is 516, for normal diet is 2262. Could anyone help
me? I will give you a big baozi if you can help me with that. Thanks a lot! |
|
q******n
发帖数: 272 | 22 use Gpower, z test of two independent proportions. you need to specify % in
normal diet group. |
|
w*****1
发帖数: 473 | 23 can I use proc power for TWOSAMPLEFREQ chi-sq test? |
|
q******n
发帖数: 272 | 24 Sure.
proc power;
twosamplefreq test=pchi
proportiondiff = .
refproportion = 0.5
groupns = 516| 2262
power = 0.9;
run; |
|
