In a clinical trial comparing two treatments, a researcher calculates the difference in means $\Delta = \mu_1 - \mu_2$. Suppose the researcher finds $\Delta$ is 12, with a variance of 4. If the researcher wants to determine if the effect is significantly greater than zero using a Wald-type test at $\alpha = 0.05$, which property of the sampling distribution allows the use of the standard normal critical value $z_{0.95} = 1.645$?