Wednesday, January 17, 2007
Independent Research
I think I want to do my project for this class on the priority pattern study on Nesticodes rufipes that I happen to be conducting for my honors thesis. Nesticodes are spiders by the way. I think looking at potential methods of analyzing data with statistics could really improve my design. Wow, and to think that this project will help me with something I already need to do!
Subscribe to:
Post Comments (Atom)
3 comments:
this sounds cool. what sort of data do you have?
Well, I actually don't have data right now. I have census numbers of my spiders and a statistical design in mind with certain minimum sample sizes. Would using sample data be an option? It might help me to get comfortable with certain types of analyses. It would also be useful for forming a more distinct hypothesis if I could think about what the data will look like. Will this be plausible at all? Should I come talk to you about my design?
using simulated data is often a really good idea to get a feel for how you'll eventually go about your analysis. i would even recommend creating two simulated (often called "dummy") data sets -- one where there is an obvious difference between groups, and one where there is none. this will make sure that your analytical procedure is adequate to test your hypothesis when you give it your real data.
as for coming to talk to me, i'm always up for that (or conversing by phone or email). but -- if you're not averse to it -- using the blog for this has the double advantage of letting other people read and get ideas from our conversation (that is, between you and me), and gives them the opportunity to offer suggestions and insights as well.
but you should do whatever you're most comfortable with. :)
Post a Comment