Are These the Same?

Abstract:Equivalence testing in product design is crucial to proving a new product works as well or better than its target or predecessors. Traditional hypothesis testing can help to prove one product is different from another, but falls short in evaluating whether two products are the same, or very close to the same. For equivalence testing, manufacturers must start with the assumption that the requirements of a new product have not been met and find evidence to prove that it has in fact met the requirements. A two one-sided test (TOST) is then performed using data on the target product specifications and customer expectations. If the two products are the same, or in most cases, close enough to not impact customer satisfaction, equivalence is …

Access this article
Other ways to access this article

Social Bookmarking

Digg, delicious, NewsVine, Furl, Google, StumbleUpon, BlogMarks, Facebook



The key here is the assumption of equivalency from experience instead of data. Just like the case study, 2.5 hours is much more than 3 sigma.
--Kyle Huang, 05-19-2017


Great article and well written. Just an minor mistake on "The robustness, as measured by their SNR, is also better for the new design (SNR = 52.71) than the standard design (SNR = 107.33)." The values in parenthesis are reversed.
--Miguel Pereda, 05-12-2017


Featured advertisers