BioGeoBEARS (version 0.2.1)

lrttest: Calculate Likelihood Ratio Test (LRT)

Description

The Likelihood Ratio Test (LRT) is a standard method for testing whether or not the data likelihood conferred by a more complex is significantly better than the data likelihood conferred by the simpler model, given a certain number of extra free parameters for the complex model. The null hypothesis is that there is no difference; rejection means that there is a statistically significant improvement in the more complex model.

Usage

lrttest(LnL_1, LnL_2, numparams1, numparams2, returnwhat = "pval")

Arguments

LnL_1

Log-likelihood of more complex model.

LnL_2

Log-likelihood of simpler complex model.

numparams1

Number of free parameters of the more complex model.

numparams2

Number of free parameters of the less complex model.

returnwhat

If "pval", just return the p-value. If "all", return all of the intermediate outputs.

Value

pval or LRT_result2. Depends on returnwhat.

Details

The LRT only works for situations in which the simpler model is nested within the more complex model (i.e., by taking some parameters of the more complex model and forcing them to be fixed to a specific value). In addition, the LRT may be unreliable in data-poor situations, and inherits whatever difficulties there may be in ML searches. See Burnham et al. (2002) for discussion.

This function assumes that LnL_1 and numparams1 refer to the more complex model, and that LnL_2 and numparams2 refer to the simpler model nested within the more complex one.

References

http://phylo.wikidot.com/matzke-2013-international-biogeography-society-poster

Burnham_Anderson_2002

Matzke_2012_IBS

See Also

lrttest_on_summary_table

Examples

test=1