Simulations and LRT test

[Miguel]

I’m using Appendix A as a guide to run a number of simulations to explore sampling effort before I begin my study. Can MARK calculate the LRT test statistic for each of the simulations? Once I open the simulation output, the tests menu is not visible. If there is a way of calculating the LRT, do I have do it for each of the individual simulations? Thank you for your help. (p.s. MARK is truly and exceptional tool!).

Miguel

[gwhite]

Miguel:
You will have to pull the simulation output DBF file into a statistical package or Excel to compute the LRT between models. This would be done for each replicate set of data, i.e., you have to run both models in each of the replicate simulations.
Gary

[egc]

Miguel:
You will have to pull the simulation output DBF file into a statistical package or Excel to compute the LRT between models. This would be done for each replicate set of data, i.e., you have to run both models in each of the replicate simulations.
Gary

The DBF data file Gary is referring to is the SIMRESULTS.DBF (name used in the example) shown at the top of p. 9 of Appendix A. The example shown in the Appendix runs both models in each of the replicate simulations (as Gary mentions), which is exactly what you want to do. What I didn't explicitly note was that you need to 'suck the DBF file' into another package to calculate summary statistics. I use SAS (which lets you import the DBF file directly), but there are a lot of things you can use (you could, for example, open up the DBF file directly in Excel, if you're more comfortable there).

I'll add a line or two to the Appendix to make this point more explicitly.