16 January 2010

Some boxplots about baby fish

It's been a while since I've said anything sciencey here because, well, being laid off from my sciencey day job has kind of drastically reduced (possibly even eliminated) the number of sciencey things I do on a daily basis. However, all is not lost! I can still dig up past sciencey things I've done, such as for my masters thesis. :)

To the right are some pictures of baby fish - five species of Lutjanus snappers that usually live in Florida but had gotten lost in North Carolina at the time they were caught. They were about 1-2 cm long at the time. We caught hundreds of them, mostly L. griseus, during two (non-consecutive) years of sampling with nets. Among other things, we compared their size ranges with each other:

These are box plots. Which is a fun modern statistical way of looking at data besides just getting an average and a standard error or standard deviation (also included, because older scientists who aren't used to fun modern ways prefer to see those there). As you can see, some of them came in bigger than others. :)

This post inspired by Janiece, who is learning about box plots in her statistics class.

2 comments:

Dr. Phil (Physics) said...

Really like how it highlights how stats suck with small data sets, like N = 2. (grin)

Dr. Phil

unsanglo = state in which you do not have any san glo, of course.

MWT said...

And don't forget the N=1 "box" either. :D