The Right Way to Detect Outliers - Outlier Labeling Rule (part 1)

[@ftothel794]

I don't think you fully comprehend HOW HELPFUL YOU AND YOUR VIDEOS ARE!!!! Thank you so much!!

[@joannaelson6166]

I don't think you fully comprehend HOW HELPFUL YOU AND YOUR VIDEOS ARE!!!! (still true)

[@imchillin2k7]

if you havent already brought out a book- please do!!!! this helped me way more than any book ive tried to decipher. thank you!!!

[@parksongkun]

Great ! Thank you for the nice solution for detecting outliers systematically !

[@salve_rex9256]

you mentioned in the video a method for non-normally distributed data. Can you give some clues?

[@nikkivanluxemburg344]

What about non normal distribution?

[@deadlypalms]

God I hope I find this video soon - my data is totally non parametric!

[@MitsosDA]

Do you have and can you test for outliers when dealing with categorical data? Thank you.

[@caaanyoudigit]

Hmm...why are my box plots and histograms blue instead of the color in the video?

[@ruth1351]

Hey thanks for the video - may I ask what if the score is minus?

[@assoc.prof.dr.hamidmohsinj5254]

Thank you very much. Is there anyone know what is the name of this method (for example: Mahalanobis , The Variance-Covariance Matrix, Fast-MCD Algorithm, .... etc)

[@dcincoltesccartofii]

if i have multiple scale scores from the same sample, what variable do i put in? the total raw score? the individual questions? do i put them all, or do i do this for every scale in particular?

[@vanessamadrazo5179]

Hello, Thanks for creating this video! :) Is there a reference citation for the Outlier Labeling Rule?

[@how2stats]

The references are in the description of the video.

[@Schatsie525]

Hey, every time I do a lineair regression, my model keeps giving me new casewise diagnotics. My KS-test is under 0.2. Can you please help me? We have already made dummies and of our independent we made a LN

[@MichelleSoekoe]

What multiplyer does SPSS use to calculate outliers? I know the norm is either 1.5 or 2.2 but I need to calculate my outliers at 3 times the IQR to the left and right of the first and third quartiles. Is there a way to change this setting in SPSS so that I can get a Box and Whiskers plot that shows the outliers with the 3* instead of what ever SPSS uses as its standard multiplyer?

[@lenka4497]

thank you so much...your videos are very helpful :)

[@humtum7983]

can you please share with us sample data

[@datsme888]

please tell which software package you used for creating your data (normal and non-normal)??

[@tabi151214]

my data has academic achievement as dependent variable and intelligence as independent....i should calculate outliers w.r.t.achievement or intelligence??

[@how2stats]

+tabi151214 Both variables need to be examined for outliers.

[@co20b]

What if my data is not normally distributed?

[@how2stats]

+Onomato Poet Consider using bootstrapping for your test statistic

[@co20b]

+how2stats How does that work?

[@lauraguerrero876]

Thanks. Very useful.

[@aminaalaqal]

thanks a lot sir it was a helpful video

[@Romeowasbleeding1]

I really really appreciate these videos. If you would like some constructive criticism, you tend to smack your lips a lot, it might be a good idea to pay attention to that while recording. Thanks once again!

[@how2stats]

What ever your doing your statistical analysis upon. If you're only going to analyse the total score, then you only need to look for outliers there. But if you plan on doing analyses at the item level, then you should look for outliers there too. I doubt you'll find outliers at the item level.

[@esunder2003]

thanks for the info.

[@speedshift2971]

who came up with the outlier formula? maybe we've been throwing out important data this whole time!