View Full Version : Statistical test, proportional increases and more confusion!


Postmodern
09-28-2008, 07:09 AM
Hey there,

I'm a Biology student and I'm finding stats quite difficult to grasp no matter how many stats101 sites I read or walk through guides I play with.

My problem is this...my experiment was as follows: varying the food source concentration available to an ant species to see if the ant traffic on the tree changed in response to this. Colonies were either subjected to a carbohydrate treatment or a protein, but these types arent comparable so seperate stats ran on each prey item. Some food patches were rich i.e. 8 mealworms or 1M sucrose, others poor i.e 1 mealworm 0.1M sucrose. We hoped that increasing quality would increase ant activity. There were 4 trails per colony, and 40 colonies per prey item type, either receiving four different mealworm abundances or 4 different sucrose abundances.

I need to work out the proportional increase/decrease of workers after the treatment was applied, but am having some difficulty doing this as I don't want a percentage, but more like 1 = baseline and 1.something would mean recruitment has been established.
I know the pretreatment would need 1 + as some of the results were 0.

Also is a GLM table just another word for an anova table?

Thanks for any light anyone can shed :)