Fail/Pass tests and the implied failure rate and confidence levels
Let's say we conduct a fail/pass test. We subject $n_\mathrm{s}$ samples to an accelerated (representative) life test of $m=1$ lifetime equivalents. The test is considered a success if $100\,\%$ of the $n_\mathrm{s}$ samples survive. Yet, this leaves the important question of how certain can we be that the population as…