Understanding Cook's Distance in SPSS

This video explains Cook’s Distance using SPSS. Cook’s Distance is a measure of influence for an observation in a linear regression. The relationship between influence, leverage, and outliers is explained.

[@FKAS8410]

no f'ing way I looked this up and its the MAN the MYTH the LEGEND doing SPSS analysis!

[@karasullivan7919]

Cook's Distance is a concept I have never been exposed to. I see the significance of such a technique to rule out weak influences.Also, I have to familiarize myself with some of the terminology.

[@SPORTSCIENCEps]

Thank you for uploading it!

[@sukhdeepchadda7463]

Very well explained Dr Todd :)

[@ElimelchRT]

Great videos, you are a good teacher, thanks for sharing, sharing is caring ;)

[@pokeefe90]

Thanks for the video. You mentioned that there are three rules of thumb when interpreting Cook's D. Could you possibly provide references for each?

[@mutindafestus5619]

this is a good video
i also have special problem of a simple linear regression
where my data has plenty of outliers
i identified the outliers +high leverage points and removed them on reruning the regression ,a new set of outliers appears
i tried this thrice and every time i remove outliers a new set emerge i am not sure how much i should remove

[@erinboardman5575]

Hi Todd,
Thanks for these videos - they have helped me immensely! You refer to the Cook's distance < 1 rule of thumb (Tabach & Fidell), but can you please advise me of a suitable ref for looking at Cook's distance relative to other cases. Thanks again.

[@hrosem]

Can you explain how you would report this finding in the results section of a report please?

[@vanessagraham6473]

Seems so simple. Will have to investigate further.

[@turgutcoskun8863]

Thank you for this nice presentation, what if we want to check the influence of more than one variable in multiple regression? Should we drag them to predictor variable box and save Cook's distance for all of them? Or, one by one ... ?

[@Btrembz54]

Thanks for the video. Do you know how to find Cooks Distance with SPSS Modeler version 18?

[@Nihonnelly]

Dear Dr. Grande, thank you very much for this video! It seems all feasible to me, yet I run a binary logistic regression in SPSS and my statistician recommended to use Cook's. I run it the way you recommended, though in the window of "Binary logistic". Yet, I cannot find in the output the cook's distance, as the table "residuals statistics" is missing. Is there another way to do that within a logistic regression? I would be very grateful if you could help me out!
Best, Neeltje

[@meysm27]

Hello Doctor
I have data for Nested ANOVA and it contains four outliers , can I use Cook distance to identify and handle outlier
also the Leverage method
Thanks

[@hopebt8961]

Thank you for your nice discussion in Cook's distance.
my question to add is what four that you divided by N(sample size)?

[@zeldafan2137]

What reference did you use for the three rules regarding Cook's distance?- to cite

[@tonyrandall3146]

You have my thanks

[@aguilarvf]

Seems simple, could you please help me understanding DFbeta and DFfits please?
Thanks

[@jet3doodleart691]

Thanks a lot. I pay $1000 for a course in my university and end up confused. I come here and clear my confusions for free.

[@Icksable]

Hi, you used 4/n to determine whether a value likely had significant influence (and thus be considered for exclusion); however you concluded that 4/n = .25. Shouldn't that have been 0.4 (i.e., 4/100 = 0.4)

[@DrGrande]

You are correct - annotation added. Thank you.

[@ningzhou4140]

0.04

[@nkristianschmidt]

6:12 4/n ... 4/100 = 0.04 ... ?

[@DrGrande]

Yes, 4/100 is .04. I added an annotation to this video. Thank you.