o******6
发帖数: 538 | 1 ☆─────────────────────────────────────☆
lucien (Jerry) 于 (Tue Aug 25 13:52:46 2009, 美东) 提到:
有若干人，每人身上抽出血样，血样分成３２份，然后分到４个TREATMENT组里，每组
８个血样（每人）。接下来，每三个月抽出一个血样（每人），看其中的有效细胞数，
这样两年后所有的８个血样（每人）全部用完。想比较４个TREATMENT中随时间变化，
哪个TREATMENT最好。
☆─────────────────────────────────────☆
DaShagen (Unbearable lightness) 于 (Tue Aug 25 13:56:44 2009, 美东) 提到:
repeated measures
☆─────────────────────────────────────☆
lucien (Jerry) 于 (Tue Aug 25 14:02:11 2009, 美东) 提到:
知道是REPEATED MEASURES， 但是样本BETWEEN | j*****e
发帖数: 182 | 2 This is a split-plot design. The whole-plot treatment is the four treatments
. It is assigned to the wholeplot unit (four blood samples from a person
under the same treatment) in a RCBD, where person is the blocking factor.
The subplot unit is each single blood sample. The subplot treatment is time.
You have to manipulate the data to a certain structure so that SAS PROC
MIXED can analyze this. Remember: design with repeated measurement is a
special case of split-plot design.
I am shocked that af | c*******o
发帖数: 8869 | 3 care to write down the SAS code for this model?
treatments
time.
【在 j*****e 的大作中提到】
: This is a split-plot design. The whole-plot treatment is the four treatments
: . It is assigned to the wholeplot unit (four blood samples from a person
: under the same treatment) in a RCBD, where person is the blocking factor.
: The subplot unit is each single blood sample. The subplot treatment is time.
: You have to manipulate the data to a certain structure so that SAS PROC
: MIXED can analyze this. Remember: design with repeated measurement is a
: special case of split-plot design.
: I am shocked that af
|
|
