I have a few sets of data and I'd like to find out what a "normal" range of values looks like in each data set. I would not like to make any assumption about an underlying distribution (so I can apply this to many data sets). I was thinking I could use an inter-quartile range analysis (box-plot with whiskers). I would get a good look at the central tendency of the data as well as an idea of how the distribution of the set looks. I can also find the outliers.
Can anyone comment on this method? What are your thoughts?