I am working with some data on hormones. The distribution is very non-normal so I log transformed the data but data is still very skewed. What can I do??
0
2 replies to this topic
#2
hobglobin
-
- Active Members
-
- 5,249 posts
Growing old is mandatory, growing up is optional...
60
Excellent
Excellent
Posted 19 February 2009 - 01:02 PM
ari_doodle, on Feb 19 2009, 09:59 PM, said:
I am working with some data on hormones. The distribution is very non-normal so I log transformed the data but data is still very skewed. What can I do??
Try out other transformations. But some data sets are skewed, nothing helps. The you should use nonparametric methods...
One must presume that long and short arguments contribute to the same end. - Epicurus
...except casandra's that belong to the funniest, most interesting and imaginative (or over-imaginative?) ones, I suppose.
#3
pcrboy
-
- Active Members
-
- 18 posts
member
0
Neutral
Neutral
Posted 20 February 2009 - 11:31 PM
try a normal score transformation, which basically will rank everything and smash everything into something with a normal distribution. the advantage of course is you get rid of outliers. the disadvantage is that you'll no longer have relative distance between each gene. if you prefer, there are nonparametric methods to analyze your data as previously suggested.
