caChiTest {InPosition}

caChiTest: correspondence analysis tests without resampling.

Description

caChiTest performs 3 sets of chi-square tests along the lines of Lebart's v-tests. These tests are designed to be conservative estimates of chi-square tests on contingency data. The tests treat this data in a standard chi-square framework, but are helpful to understand correspondence analysis data when permutation and bootstrap become unfeasible.

Usage

caChiTest(DATA, res, critical.value = 2)

Arguments

DATA

Data as would be entered for Correspondence Analysis (see link{epCA})

res

Results from correspondence analysis (e.g., output from link{epCA}).

critical.value

numeric. A value, analogous to a z- or t-score to be used to determine significance (via bootstrap ratio).

Value

a list with the following values:

j.sig.vals

boolean matrix. Identifies which column items are significant (based on critical.value).

j.signed.vals

chi-square values associated to column items, multiplied by the sign of their component scores ($fj).

j.p.vals

p values associated to column items in a chi-square test.

i.sig.vals

boolean matrix. Identifies which row items are significant (based on critical.value).

i.signed.vals

chi-square values associated to row items, multiplied by the sign of their component scores ($fi).

i.p.vals

p values associated to row items in a chi-square test.

omni.val

chi-square value associated to the table.

omni.p

p value associated to a chi-square tests of the table.

Author(s)

Derek Beaton

See Also

epCA.inference.battery