c*******o
发帖数: 8869 | 1 我有0，3，8，24 四个药物dose, 有没有什么trend test 测定剂量和反应之间的关系
？ 我这里指的不是linear regression, 毕竟四个剂量太少了， 不足以做continuous
variable |
D******n
发帖数: 2836 | 2 4 is not too few ah
continuous
【在 c*******o 的大作中提到】
: 我有0，3，8，24 四个药物dose, 有没有什么trend test 测定剂量和反应之间的关系
: ？ 我这里指的不是linear regression, 毕竟四个剂量太少了， 不足以做continuous
: variable
|
c*******o
发帖数: 8869 | 3 then you think linear regression is OK? I just vaguely recalled something
like more formal "trend test" exist which I may want to try out....
【在 D******n 的大作中提到】
: 4 is not too few ah
:
: continuous
|
D******n
发帖数: 2836 | 4 there is sth called polynomial orthoganl ,sth like that to test linear trend
for ordinal categorical predictor variables, i am not sure about the name n
ow.
【在 c*******o 的大作中提到】
: then you think linear regression is OK? I just vaguely recalled something
: like more formal "trend test" exist which I may want to try out....
|
s*r
发帖数: 2757 | 5 whether you can use linear regression depends on the distribution of the
endpoint (conditional upon X)
if your endpoint is binary, google armitage trend
continuous
【在 c*******o 的大作中提到】
: 我有0，3，8，24 四个药物dose, 有没有什么trend test 测定剂量和反应之间的关系
: ？ 我这里指的不是linear regression, 毕竟四个剂量太少了， 不足以做continuous
: variable
|
c*******o
发帖数: 8869 | 6 response variable is continuous. in fact, I already have p value of these
end points for the comparison between each dose and control level (i first
treat dose as categorical variable). i am just not sure the reliability of
these results because each dose level only has two replicates! think about,
this is essentially 2 vs 2 comparison althought the error term is estimated
from the whole sample. I need some kind of trend test (i heard this term
before and pretty sure they do exist) to assure the
【在 s*r 的大作中提到】
: whether you can use linear regression depends on the distribution of the
: endpoint (conditional upon X)
: if your endpoint is binary, google armitage trend
:
: continuous
|
s*******r
发帖数: 181 | 7 To me, your term "trend test" is totally same as simple linear regression. |
s*r
发帖数: 2757 | 8 you are looking for the thing mentioned by DaShagen
google "trend test proc glm contrast"
,
estimated
endpoints
【在 c*******o 的大作中提到】
: response variable is continuous. in fact, I already have p value of these
: end points for the comparison between each dose and control level (i first
: treat dose as categorical variable). i am just not sure the reliability of
: these results because each dose level only has two replicates! think about,
: this is essentially 2 vs 2 comparison althought the error term is estimated
: from the whole sample. I need some kind of trend test (i heard this term
: before and pretty sure they do exist) to assure the
|
