The Student’s t-test is rearranged in order to predict the sampling (N) required to conclude significant difference between two observation sets at a particular confidence level. The expression is appropriate for small or large observation sets. When the difference in means is small relative to the standard deviation, more tests are required in order to reliably detect that difference.
Introduction
In 2012, Keysight Technologies, Inc. introduced the Express Test option for the G200 NanoIndenter platform. This option implements traditional indentation testing in a revolutionary way in order to achieve unprecedented testing speeds[1,2]. Express Test performs one complete indentation cycle per second, including approach, contact detection, load, unload, and movement to the next indentation site. One hundred indentations can be performed at one hundred different sites in less than 100 seconds. Given that indentations can be performed so quickly, a new question arises: How much testing is enough testing? The purpose of this note is to answer this question by applying the Student’s t-test in an uncommon way.