0

I have two computers that have access to the same sensor on a plane and both log speed every second. What sample size formula should I use for testing a null hypothesis of equivalence in terms of speed logged?

Assuming the speeds logged are normally distributed

#1 Sample Sizes for Two Independent Samples, Continuous Outcome

#2 Sample Size for Paired Samples, Continuous Outcome

Pitouille
  • 1,506
  • 3
  • 5
  • 16
Data plumber
  • 11
  • 2
  • When you say "what sample size formula should I use for testing a null hypothesis of equivalence in terms of speed logged", I assume that you are looking for a statistics test that allows you to determine whether both computers are logging speed the same way? – Pitouille Sep 21 '21 at 15:41

1 Answers1

0

Too few information about your problem is provided. My understanding is that you want to perform an equivalence test. The simplest approach is the TOST (Two-One-Sided T-test) procedure https://en.wikipedia.org/wiki/Equivalence_test. It is already well documented on this site https://stats.stackexchange.com/a/500121/321901

Concerning the sample size, same comment, too few information is provided, but you might be interested by this article: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5012670/ and the R package powerTOST

Pitouille
  • 1,506
  • 3
  • 5
  • 16