Linear Transformation Problem

#1
Hi there,

Basically, I am checking for outliers in a 'frequency of use' construct measured in number of weeks. Since the box plot was right-skewed, i conducted a log transformation via spss. Although, the result seemed fine but now I want to create boxplots for other constructs which are measured on 7-point likert scales however, 'frequency of use' is not. My question is how do I perform a linear tranformation (preferably in SPSS) on the new 'frequency of use' scale so that it can be measured on a 7-point likert scale like the others when doing the box plots?

Please let me know if any part of my question is unclear.

Thanks
 

noetsi

No cake for spunky
#2
A minor aside. The boxplot is never right skewed. The data it analyzes may be.:p There is a special form of boxplot in R that adjusts for skewed data and if you have access to R you might try that.

I don't understand why a transformation changed the number of categories in your likert scale for "frequency of use." If that indeed is what you mean. I also don't understand why it matters how many levels your variables have in generating outliers. Outliers in a box plot are tied to comparisons to the median value (or more accurately the IQR) not how many levels a given variable has.