English Deutsch Français Italiano Español Português 繁體中文 Bahasa Indonesia Tiếng Việt ภาษาไทย
All categories

I am looking at # of individuals in a bird study in 4 different habitats. My data reads like:
Location #1
Red-tailed hawk - 2
some gull - 4
canadian geese -7
etc.

And i have 3 other locations to compare it to. What stat test should I use?

2006-11-29 17:20:19 · 1 answers · asked by An Agent of Chaos 5 in Environment

1 answers

It depends on what hypothesis you are testing - differences in number of species among sites, differences in number of individuals among sites, differences or patterns in the number of species and individuals, or maybe you want to actually compare the habitat types based on the species and numbers.

I am guessing that you want to look at similarities or differences in the habitats, based on the species and their numbers. One way to do this is with the Kolmogorov-Smirnov (KS) test which will tell you which habitats have significantly different distributions of bird species - to do this all of the sites must have the same list of species in the same order, so you need to put 0's in for those species that were not found at a particular site; and all of the sites need to be compared either to the mean or grand tally of species and numbers from all sites.

Another way is to use cluster analysis, which is a way to characterize similarities or differences among the habitats. Of these approaches, only the KS test can easily be done with a spreadsheet like Excel, while cluster analysis must be done with matrix algebra or a statistics program..

2006-12-02 09:36:25 · answer #1 · answered by formerly_bob 7 · 0 0

fedest.com, questions and answers