+ Reply to Thread
Page 2 of 2 FirstFirst 1 2
Results 16 to 16 of 16

Thread: Removing an outlier DECREASES correlation??

  1. #16
    Points: 1,721, Level: 24
    Level completed: 21%, Points required for next Level: 79

    Thanked 0 Times in 0 Posts

    Just looking back over the thread it seems there are a few causes for concern:

    1) Just looking at the data - your correlation is not robust at all. You do have a very serious outlier and you would need to explain how it came to be there in before you decided to keep it - why is this data point so high when the others are low. The most interesting thing about your data is that one point - why? With it in the model you can not run the usual tests for significance - because it violates all the assumptions.
    2) Further since you have a small sample size, the assumptions of the general linear model are critical for any meaningful interpretation of the results.

    I just ran some diagnostics on your data - attached. Hope this helps. Good luck sleuthing the origin of that crazy number.
    Attached Images

+ Reply to Thread
Page 2 of 2 FirstFirst 1 2


Similar Threads

  1. Outlier
    By Hubert12 in forum Statistics
    Replies: 6
    Last Post: 09-29-2014, 12:31 AM
  2. Removing insignificant variables
    By constantly in forum Regression Analysis
    Replies: 2
    Last Post: 03-21-2011, 01:49 AM
  3. LM test for removing parameter
    By thedreamshaper in forum R
    Replies: 0
    Last Post: 12-01-2010, 02:33 PM
  4. Outlier
    By iced_eax in forum Biostatistics
    Replies: 0
    Last Post: 07-25-2010, 07:26 PM
  5. removing blanks across columns
    By shabani in forum SAS
    Replies: 0
    Last Post: 06-01-2009, 04:08 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts

Advertise on Talk Stats