Dear R-lister and Prof. Harrell,
To Prof. Harrell: I feel very sorry to bothering you.
The ecological data usually do not follow strictly-defined normal
distribution. So, there are many more works before any parametric testing or
any regression modeling-fitting .
Tens of days ago, when I ask for proper methodology of outliers treatment in
the data processing, Prof. Harrell ever adviced me to turn to nonparametric
test. Since then, I put my efforts on setting up of my nonparametric test
techniques. But, still now I can not gain any progresses on that.
I was trained as evolutionary biologist (with background of genetics and
ecology), without solid statistical trainnings, especially on nonparametric
test techniques. And, I was educated in China where it is very hard to reach
the statistician's supports. In the last five year, I have builded up a
perfect (in my view) dataset which integrated genotype assignment, more 50
ecological and morphological traits, georeference of each sample (more than
1000 in total), and various of environmental parameter (gained by GIS
searching in geoinformatic databases).
Although I have much experiences on parametric statistics, I have not
ultilized any noparametric statistics on my research except on some
black-box softwares.
I am thirst for any advice on non-parametric statistics on ecological data
analysis. Is there any example-oriented books on that? How it can be
implemented in R?
Best regards,
Yours
Mao J-F
[[alternative HTML version deleted]]
