Monetary Unit Sample - Poisson

Comments

• koenderks

  November 2021

  Hi Johan,

  It seems to me like you want to be able to state after your sample that the misstatement in the population is lower than 5,000,000 (10 percent of the population size/value). If you want to make this statement with a certain amount of confidence (say 90%) you will need at least 24 samples provided that you tolerate zero (0%) errors in these 24 samples. The expected errors setting determines the number of tolerable errors k in the sample of n items.

  
  

  Planning for zero errors is efficient but the assumption that you find zero errors in the sample might not hold. Therefore you can adjust the number or percentage of expected errors to increase your sample size, mitigating the risk that you find more errors than you have planned for initially, which would imply that insufficient work has been done in the end. By setting the expected errors to absolute and choosing 1, you now tolerate 1 error in the sample and will require a larger sample size.

  by setting the expected errors to relative, (5%) for example, you set k as a fraction of n. This means that you expect 5% of your sample to contain an error, which in the case of 80 samples amounts to 80x0,05=4 errors.

  Hope this clarifies.

  Two suggestions:

  • you may download an updated version of the audit module in JASP 0.16 (via www.jasp-stats.org)
  • you can find the manual of the audit module here: https://github.com/jasp-stats/jaspAudit/raw/master/man/manual.pdf, here also more information about the High, Medium, Low settings can be found.
  • every analysis also has their own help files in which most of the options are explained. You can find these help files by clicking on the “i” icon In the interface next to the title of the analysis.

  Best,

  Koen
• EJ

  November 2021

  I'll forward this to Koen, our expert for this analysis

  E.J.