Hi
I have a sample of 250+ cases, 7 variables and each variable have items ranging from 4 to 15. I detected the univariate outliers.
I want to detect the Bivariate outliers and multivariate outliers using SPSS. So my question is for the bivariate outliers should I compute the items in each variable in a total variable first and then conduct the bivariate outliers detection as a second step? or how this is done ??
Any helpful resources
Please share any useful steps regarding this.
Thanks in advance
Nada Rabie
Here are a couple of methods I like which are somewhat related to Ali H Abuzaid suggestion. I like these because I think using robust methods make things a little more accurate in the presence of outliers. Best, David E. Booth
I recommend performing the mahalanobis distance to detect atypical data and improve the distraction of normality.
https://www.youtube.com/watch?v=AXLAX6r5JgE
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS