https://community.jmp.com/t5/Discussions/Standardization-required-in-Multivariate/m-p/768154#M94838 <P>Here are my thoughts:</P> <P>&nbsp;</P> <P>Multivariate is typically done on the Y's (response variables), not the x's. &nbsp;Now that doesn't mean you can't do both.</P> <P>The way Mahalanobis works is to relate the values in each column to the values in other columns over the data set. &nbsp;The analysis evaluates, in general, are you getting values reasonably close enough to typical values when the other columns are their respective levels. &nbsp;If not, perhaps something unusual happened during that <EM>row</EM>. &nbsp;It is not specific as to which data points created the infraction. &nbsp;Sometimes it's obvious which value was unusual, sometimes not.</P> <P>&nbsp;</P> <P>In any case, standardization of the values is not necessary.</P> Mon, 24 Jun 2024 22:38:37 GMT statman 2024-06-24T22:38:37Z https://community.jmp.com/t5/Discussions/Standardization-required-in-Multivariate/m-p/768144#M94837 <P>I'm trying to perform an outlier analysis (Mahalanobis distance) in Multivariate. Do I need to standardize (normalize all values to the same scale, for example, -1 to 1) all values before executing the analysis, or does JMP do that for me automatically? Thanks.</P> Mon, 24 Jun 2024 22:31:02 GMT https://community.jmp.com/t5/Discussions/Standardization-required-in-Multivariate/m-p/768144#M94837 JZ 2024-06-24T22:31:02Z https://community.jmp.com/t5/Discussions/Standardization-required-in-Multivariate/m-p/768154#M94838 <P>Here are my thoughts:</P> <P>&nbsp;</P> <P>Multivariate is typically done on the Y's (response variables), not the x's. &nbsp;Now that doesn't mean you can't do both.</P> <P>The way Mahalanobis works is to relate the values in each column to the values in other columns over the data set. &nbsp;The analysis evaluates, in general, are you getting values reasonably close enough to typical values when the other columns are their respective levels. &nbsp;If not, perhaps something unusual happened during that <EM>row</EM>. &nbsp;It is not specific as to which data points created the infraction. &nbsp;Sometimes it's obvious which value was unusual, sometimes not.</P> <P>&nbsp;</P> <P>In any case, standardization of the values is not necessary.</P> Mon, 24 Jun 2024 22:38:37 GMT https://community.jmp.com/t5/Discussions/Standardization-required-in-Multivariate/m-p/768154#M94838 statman 2024-06-24T22:38:37Z