Statistics dialemma?

Question

Hi Folks, I'm just wondering which statistical measure would be appropriate for the following data. In short, I am doing a project that compares the timing of speciation events between plants and animals. My data will look something like this (just an example...)

Plants:

Daisy 2 - 5 Million years ago (Mya)

Pansy 45 - 56 Mya

Rose 12 -18 Mya

Holly 7 -10 Mya

Animals:

Tiger 2-4 Mya

Toad 34 - 39 Mya

Lemur 3-4 Mya

Plants have about 30 data points and animals have about 50. I am comparing the ranges of dates of plants and animals and seeing if there is any statistical difference between them, If you could point me in the direction of which statistical test would be best for this kind of data I will be eternally grateful to you! Sorry if I haven't explained myself very well x*x

Answer 1

I am not a statitician, but I think there are several different questions you can ask of such data and that then the test would be different.

1) did plants arrive earlier than animals. If the results are normally distrbuted you could use a "t-test". If not then use a non-parametric test such as Wilcoxian rank sum (Mann Whitney-U test).

However my concern with this would be how you ensure your data is not biased, so I have mostly the more "recent" mammals  (as your list suggests) or the "older" amphibians or even the celocanth (80+ million years ago). Don't have a good idea about how you ensure this isn't a problem.

2) are plants appearing more often than animals - this would invole comparing the spread of the data.  This would be the comparison of the variance/standard deviations for each group.