What is the difference between the null and alternative hypotheses statements in one-tailed and two-tailed tests? How can manufacturing companies use the standard normal distribution to determine quality control of their products?
One-tailed tests only work in one direction, while two-tailed tests work in both directions.
One-tailed
Ho: A = B
H1: A < B or A > B (not both)
Two-tailed
Ho: A = B
H1: A ≠ B
I would think the companies would not want their products to deviate significantly from the standard in either direction.