Thread:Feedback from trial session (2)

Thanks Alison,

I just found your comments today. I remember why I have a hard time using the discussion threads....

I appreciate the input. I love the formulae you came up with. I'm surprised by the variability in the measurements. Makes me wonder if this approach is viable. I'll put a bunch of students onto this in my Spring course and see what sort of variability they come up with along with a comparison to original measurements.

The photos are about 3,000 pixels wide. I agree that greater precision can be achieved with a larger image. If the approach has a bias, there is no reason to expect the bias would create fake patterns...rather I'd expect either a consistent bias in one direction affecting all skulls, or increasing variance that would tend to obscure real patterns. My logic is that this is a conservative approach...if we find a pattern it's probably real.

In Spring I'll get 2 things going: more skulls; more measurements. Please thank your students for all their work. If they are interested in environmental data please let me know....land use..water quality....insect diversity. I have lots: http://www.uvm.edu/~streams/

Cheers, Declan