compRejectionFraction: Compute rejection fraction for sequential tests.
In PsumtSim: Simulations of grouped responses relative to baseline.

Description Usage Arguments Details Value Note Author(s) References Examples

Computes the fraction of simulated neurons with Poisson spike counts which are rejected by a sequence of tests. First a t-test for a difference from baseline for any category, followed by an ANOVA of an effect of category on the cells found to be significant in the first test.

1 2	compRejectionFraction(bkgLevel, respLevel, numCats, pretestP, anovaP, showProgress = FALSE, numTrialsPerCat = 10, numCells = 1000)

`bkgLevel`	Average firing rate for background counts.
`respLevel`	Average firing rate for response counts.
`numCats`	Number of categories to test for a response in.
`pretestP`	p-value to use in first t-test for a difference from baseline.
`anovaP`	p-value to use in second ANOVA testing for an effect of category on the responses.
`showProgress`	TRUE if should list cell number as calculating. Default is FALSE.
`numTrialsPerCat`	Number of trials, with background and response counts, in each category. Default is 10.
`numCells`	Number of cells to simulate. Default is 1000.

If the first and second tests were operating independently, the rejectionFrac would remain constant and equal to the anovaP value for all exclusion fractions.

`exclusionFrac`	Fraction of cells which were not rejected by the first t-test; thus fraction of those ignored for the second test.
`rejectionFrac`	Fraction of cells rejected by the first test which were rejected by the second test.

Both the t-test and the ANOVA assume a normal distribution of the counts.

Peter N. Steinmetz <Peter.Steinmetz@steinmetz.org>

Steinmetz & Thorp, 2012.