KM.calc.sample: Calculate sample size for KM-audit.

In elec: Collection of Functions for Statistical Election Audits

Calculate sample size for KM-audit.

Description

Calculate the size of a sample needed to certify a correct election if a KM audit is planned.

Usage

KM.calc.sample(Z, beta = 0.75, taint = 0, bound = c("e.plus", "WPM", "passed"))

Arguments

Z	elec.data object
beta	Desired level of confidence. This is 1-risk, where risk is the maximum chance of not going to a full recount if the results are wrong. Note that in Stark's papers, the value of interest is typically risk, denoted $alpha$.
taint	Assumed taint. Taint is assumed to be the taint for all batches (very conservative). If taint=0 then we produce a good baseline.
bound	Type of bound on the maximum error one could find in a batch.

Value

A audit.plan.KM object.

Author(s)

Based on the KM audit by Stark.

See Also

KM.audit

Examples

  data(santa.cruz)
  Z = elec.data( santa.cruz, C.names=c("danner","leopold") )
  KM.calc.sample( Z, beta=0.75, taint=0 )

elec documentation built on April 27, 2022, 1:05 a.m.