How to analyse data


I have data on real locations of dead animals and paired random locations to simulate where that animal could have died in it's home range all being equal. I have done chi-squared testes to compare the differences in e.g. habitat of dead animal between real and random locations.

1. Is this valid?
2. is there a model I can run to get better information?

I had thought about BLR but that will also try and predict the random locations, I need some way to separate this out.

Unfortunately, I don't have data on e.g. % of habitat in the area so I cannot look at the chances of that real death being in that habitat type based on make up of the area hence the creation of random points and looking to compare those.

I had thought about seeing if there was a model to use based on positive only results (i.e. real only locations), however, such a model wouldn't let me know if the death was in woodland because there is 95% woodland inn the area and therefore a greater chance of it occurring there or only 15% woodland but it increases death rate for some reason.

Thanks in advance!