topic Re: analysis on highly skewed data in Statistical Procedures https://communities.sas.com/t5/Statistical-Procedures/analysis-on-highly-skewed-data/m-p/151053#M7949 <HTML><HEAD></HEAD><BODY><P>ideally the geometric mean, if that makes sense.</P><P>Thank you for your reply!</P></BODY></HTML> Tue, 03 Feb 2015 17:34:40 GMT SRaehsler 2015-02-03T17:34:40Z analysis on highly skewed data https://communities.sas.com/t5/Statistical-Procedures/analysis-on-highly-skewed-data/m-p/151051#M7947 <HTML><HEAD></HEAD><BODY><P>Hello, as a recent graduate I need some help using SAS. I have 3 groups which I would like to compare to see if they differ among 4 variables. Group 1 (n=42) has skewness values of around 3. Group 2 (n=101) has a skewness of around 5. The last group (n=10,106) has skewness values between 8 and 10. Like I said, I would like to determine if there are any statistically significant differences between the 3 independent groups for each of the 4 different continuous variables. I am working with NHANES data, since it is a cross-sectional study I BELIEVE that means I must use nonparametric tests? What are your suggestions as to how I should complete this analysis using SAS. Someone has suggested I compare the geometric means, but I do not know enough about geometric means to know if this is the best method.</P></BODY></HTML> Tue, 03 Feb 2015 16:26:36 GMT https://communities.sas.com/t5/Statistical-Procedures/analysis-on-highly-skewed-data/m-p/151051#M7947 SRaehsler 2015-02-03T16:26:36Z Re: analysis on highly skewed data https://communities.sas.com/t5/Statistical-Procedures/analysis-on-highly-skewed-data/m-p/151052#M7948 <HTML><HEAD></HEAD><BODY><P>Statistically significant differences of what: distribution, mean, variance or something(s) else?</P></BODY></HTML> Tue, 03 Feb 2015 17:30:13 GMT https://communities.sas.com/t5/Statistical-Procedures/analysis-on-highly-skewed-data/m-p/151052#M7948 ballardw 2015-02-03T17:30:13Z Re: analysis on highly skewed data https://communities.sas.com/t5/Statistical-Procedures/analysis-on-highly-skewed-data/m-p/151053#M7949 <HTML><HEAD></HEAD><BODY><P>ideally the geometric mean, if that makes sense.</P><P>Thank you for your reply!</P></BODY></HTML> Tue, 03 Feb 2015 17:34:40 GMT https://communities.sas.com/t5/Statistical-Procedures/analysis-on-highly-skewed-data/m-p/151053#M7949 SRaehsler 2015-02-03T17:34:40Z