K2K
★    

India,
2013-02-20 09:10
(4055 d 05:04 ago)

Posting: # 10075
Views: 6,391
 

 Cook distance [Outliers]

Dear All

can any one provide me the literature for Cook distance and Table for Critical Value reference.

Thanks in advance

thanks and regards
Karthik
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2013-02-20 15:49
(4054 d 22:25 ago)

@ K2K
Posting: # 10078
Views: 5,287
 

 Cook’s distance

Hi Karthik,

see Wikipedia.
Run the analysis on the model’s residuals. I would not try to dive into critical values because many approaches have been suggested in the literature. Simply draw an xy-plot, where x is the subject’s index and y Cook’s distance. If there are “influential values” they will catch your eye. ;-)

Dif-tor heh smusma 🖖🏼 Довге життя Україна! [image]
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes
K2K
★    

India,
2013-02-21 07:40
(4054 d 06:34 ago)

@ Helmut
Posting: # 10090
Views: 5,057
 

 Cook’s distance

Hi helmut,

❝ Simply draw an xy-plot, where x is the subject’s index and y Cook’s distance. If there are “influential values” they will catch your eye. ;-)


[image]by using the prog reg and gplot (subject vs dffits) I plot the graph for cook distance method, whether its correct.



Thanks for your reply.

thanks and regards
Karthik
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2013-02-21 18:44
(4053 d 19:30 ago)

@ K2K
Posting: # 10095
Views: 5,106
 

 Cook’s distance

Hi Karthik,

❝ by using the prog reg and gplot (subject vs dffits) I plot the graph for cook distance method, whether its correct.


I don’t know this software. Looks like you have included both treatment periods from a 2×2×2 cross over study. Residuals from both periods should have the same values and differ in their signs. Rerun the analysis with one period only. Here is an example obtained with R, package bear:
[image]

Dif-tor heh smusma 🖖🏼 Довге життя Україна! [image]
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes
K2K
★    

India,
2013-02-22 13:22
(4053 d 00:52 ago)

(edited by K2K on 2013-02-22 22:46)
@ Helmut
Posting: # 10103
Views: 4,888
 

 Cook’s distance

Hi Helmut,

thanks for your reply and suggestion.

thanks and regards
Karthik
UA Flag
Activity
 Admin contact
22,957 posts in 4,819 threads, 1,636 registered users;
127 visitors (0 registered, 127 guests [including 3 identified bots]).
Forum time: 14:15 CET (Europe/Vienna)

With four parameters I can fit an elephant,
and with five I can make him wiggle his trunk.    John von Neumann

The Bioequivalence and Bioavailability Forum is hosted by
BEBAC Ing. Helmut Schütz
HTML5