Covrg {CooccurrenceAffinity}

Coverage Probabilities for Confidence Intervals about alpha, for fixed true alpha

Description

This function calculates the coverage probability at the true value alpha of the four types of Cpnfidence Intervals (CI.CP, CI.Blaker, CI.midQ, CI.midP) computed in AlphInts().

Usage

Covrg(marg, alph, scal = log(2 * marg[3]^2), lev = 0.95)

Arguments

marg

a 3-entry integer vector (mA,mB,N) consisting of the first row and column totals and the table total for a 2x2 contingency table

alph

True log-odds-ratio value alpha at which coverage probabilities (under Extended Hypergeometric with parameters mA,mB,N, exp(alp)) are to be calculated

scal

an integer parameter (default 2*N^2, capped at 10 within the function) that should be 2 or greater

lev

a confidence level, generally somewhere from 0.8 to 0.95 (default 0.95)

Details

See AlphInts() documentation for details of computation of the four confidence intervals CI.CP, CI.Blaker, CI.midQ, CI.midP. The confidence intervals are calculated for each x in the allowed range from max(mA+mB-N,0) to min(mA,mB), and the probability that X=x times the indicator of alph falling in each of them is summed.

Value

A vector covPrb containing the coverage propbabilities for the four Confidence Intervals

Author(s)

Eric Slud

References

to be added

Examples

Covrg(c(50,70,150), 1.2, lev=0.95)
Covrg(c(50,70,150), 0, lev=0.95)
Covrg(c(50,80,120), 1.5, lev=0.9)