1. Start with some desired level of confidence: C.
2. C is the central area under the normal curve that we want to catch.
3. Now, we want to find z*, the number sor that any normal distribution has probability C within z* and –z* standard deviations of its mean.