TABLE OF CONTENTS To determine whether the same product from two different manufacturers is reliabilitywise different, or a redesign has improved the reliability of a product, or a process change has improved the reliability of a component, do the following:
Take one sample each from two manufacturers of the same product, or from the original and the new design, or from the previous and the new process.
Conduct binomial (success and failure type tests of the Bernoulli trials kind) tests on these two samples and determine the successes and the failures.
Fill in Table 2.1.
Calculate the observed 
value from
| (2.1) |  |
Choose a level of significance, ?, enter Table 2.2 and find the critical value of the ? 2, at this level of significance, 
.
If
conclude that a difference does exist at the 100 ?% significance level; or with only a 100 ?% risk, we may incorrectly conclude that there is a difference when, in fact, they are equal or come from the same population.
If the condition
prevails, then conclude that there is no difference at the 100 ?% significance level and the observed difference is due to chance alone, which is to be expected because of the inherent variability designed, built and tested into the product and which could be exhibited by the two samples were they drawn from the same population (the same product coming off the same production...
