CURM Final Analysis....

From Norsemathology
Jump to navigation Jump to search


To finalize our plans for this first paper for the Florida Epidemiologist, and to cover Grayson and Katie for their final CURM responsibility.

To Dos

R Development Core Team (2008). R: A language and environment for
 statistical computing. R Foundation for Statistical Computing,
 Vienna, Austria. ISBN 3-900051-07-0, URL

  • Katie:
    • Redoing some figures, to
      • include larger font legend information
      • include special x axes values at which things are going on (stair-step method)
      • and add N -- sample size -- to the results
    • Add new figure on linear models of the cicada mass/RWL relationship.
    • Tone down regression results
  • Grayson:
    • Write a letter describing our work for CURM, to accompany the "final" copy of the manuscript. This should be a joint letter from you and from Katie, detailing (albeit briefly):
      • The presentations:
        • Departmental Presentation
        • CURM presentation
        • MAA section meeting
        • NKU Celebration of Student research
      • The Resources:
      • The Manuscript
        • Jon's "final" for the Florida Entomologist
      • A note that more of our findings should be released in a follow-up paper, and that we'll be sure to keep him and CURM in the loop.
      • Finally you should probably thank him for giving us this opportunity to work on an interesting project and some interesting mathematics. I hope that you've had fun, and have learned a few things, and that you will convey that to him.
    • Calculate sweet spot values (are you getting 2.09 and 1.90?)
    • Can we do the regression of the cumulatives in R? That would be better.
  • All:
    • I mused about whether the wasp RWL/mass relationship had slope significantly different from 3. It does!
    • Think on these thoughts of Jon's:

Just some ramblings:

I plan to spend much more time reading and thinking about pressures of natural selection that influence body size in provisioning wasps. The sexual size dimorphism ratio (sample mean female wet mass/sample mean male wet mass) is smaller for the Newberry wasps (1.86) than it is for the St. Johns wasps (2.14). Does this indicate that large size in females is more advantageous in the latter than in the former.

Andy responds:

This could be a consequence of the 4/7 split on small cicadas, versus
the sweet spot 1/2 split of St. Johns. In fact,
is almost exactly equal to
2/1 / 7/4

The Newberry females seem to be very inefficient; why not cut way back on the provisions given to daughters and make them just large enough to carry the small cicadas? Why not produce daughters that are only 1/2 the size of the small cicadas? In other words, why not give daughters just 2 little cicadas?

Andy responds:

Would they actually develop on that much? Or would they be stunted or starved to death?

It doesn't make sense to produce daughters that are larger than small cicadas yet too small to provision with medium or large cicadas; it is wasted reproductive effort.

Are females somehow programmed to reproduce daughters that are the same size that they are?

Andy responds:

This is certainly something that we've conjectured: that the female is
"reproducing herself (and her mate)" by creating nest cells that are proportional
to her body size -- then filling them half full, or completely full. 

I welcome all ideas you, Katie, and Grayson have about this particular size issue.


Final Analysis

Jon Hastings's rough draft for this manuscript, destined for the Florida Entomologist

Now we want to complete our Final Analysis Section for the Florida Epidemiologist

Data sets

We've verified that we're using the proper data sets (conforming to those that Jon wants to use) in the following: