catlearn (version 0.8)

nosof88oat: Ordinal adequacy test for simulations of nosof88 CIRP

Description

Tests whether a model output passes the ordinal adequacy criteria for the nosof88 CIRP.

Usage

nosof88oat(dta, xtdo=FALSE)

Arguments

dta

Matrix containing model output. The matrix must have the same format, row order, and column names, as data(nosof88); with that proviso, the output of any simulation implementation can be handled by this function.

xtdo

eXTenDed Output: Either TRUE or FALSE

Value

Where xtdo=FALSE, this function returns TRUE if the ordinal adequacy tests are passed, and FALSE otherwise.

Where xtdo=TRUE, this function returns a summary matrix. The columns are stimulus numbers. The rows ('B','E') indicate the baseline (equal frequency) condition ('B') and the experimental conditions ('E2' or 'E7', depending on the column).

Details

This function implements the Wills & O'Connell (n.d.) ordinal adequacy tests for the nosof88 CIRP. Specifically, a model passes this test if it passes all four component tests: 1. E2(2) > B(2), 2. E7(7) > B(7), 3. E2(4) > B(4), 4. E7(9) > B(9). These tests refer to classification accuracy for particular stimuli in particular experimental conditions. For example, E7(9) indicates stimulus 9 in experimental condition E7.

Alternatively, by setting xtdo to TRUE, this function returns the summary model predictions reported by Wills & O'Connell (2016).

References

Wills, A.J. & O'Connell (n.d.). Averaging abstractions. Manuscript in preparation.

See Also

nosof88