Prob(|Z|>=1.96) <= 0.04999565
This is not a test, but a distribution. The Normal distribution, or Gaussian distribution, is the statistical distribution. It's importance flows from the fact that :
Z = (x-mean)/(standard deviation)
and has a mean value of zero and a standard deviation of one.
The distribution of Z has mean 0.
Z has a Standard Normal distribution with mean = 0 and standard deviation = 1.
This "test" can only be used when there is no reason to doubt this assumption.
If the mean and standard deviation are not 0 and 1 respectively, calculate
Z = (x-mean)/(standard deviation).
If x is the mean of a sample of size N, use (population standard deviation)/sqrt(N) instead of (population standard deviation).
Be aware that both the mean and the standard deviation must be known a priori and should not be estimated from the sample itself (use a Student t-distribution if you do not know the mean and standard deviation).
Level of significance:
Use a table to look up the level of significance associated with Z.
Most used "test" because it is the approximation of all other tests.
The example used here approximates the Standard Normal distribution with an error < 1.5 * 10(-7) ( Abramowitz and Stegun 1965 (1970), p932, 26.2.17)