Focus on
Sampling used as substantive tests ($ balance testing)
Variable
Probability Proportionate to Size Sampling (no std dev, easy to design, smaller sample)
Relationships to Sample Size:
Direct = Expected misstatement, Standard Deviation, Assessed lvl of risk (rmm), Variance
Inverse = Tolerable misstatement, Acceptable lvl of risk (audit risk, incorrect acceptance/rejection)
Variable sampling plans to estimate true population value
Mean per unit (MPU) = avg sample value $ x # of items in population – Does NOT req book value (BV)
Ratio Estimation = (Audited TV/Audited BV) x Total BV of population
Difference Estimation = ((Audited BV – Audited TV) / sample size) x Population
Attirbute Sampling (Testing of controls – existence of error needs to be verified or debunked – yes/no questions)
Req. audit trail to be useful
Relationships to sample size
Direct = Expected error rate (deviation means error)
Inverse = Risk of assessing control risk to low, Tolerable deviation rate
Evaluating Sample Results
in order to rely on controls
Sample Error Rate + Allowance for Sampling Risk < = Tolerable deviation
If SER + AER exceeds TER than do not rely on controls and modify the NET of substantive tests