What is the difference between one-tailed and two-tailed tests?

• A. One-tailed tests test both directions; two-tailed tests test a specific direction.
• B. One-tailed tests have a two-sided p-value.
• C. One-tailed tests test a specific direction; two-tailed tests test both directions.
• D. One-tailed tests use alpha; two-tailed tests do not.

Answer: (C)

Direction of the effect is what distinguishes one-tailed from two-tailed tests. A one-tailed test uses a hypothesis that specifies the direction of the effect, such as a treatment increasing a measure above a standard. You reject the null only if the observed statistic falls into the extreme region in that single direction, and the p-value reflects the probability of observing something as extreme in that one direction.

A two-tailed test asks whether the mean differs from the hypothesized value in either direction. The alternative is μ ≠ μ0, so the rejection region lies in both tails of the distribution, and the p-value accounts for extremes in either direction (effectively doubling the one-sided tail probability).

The practical takeaway is that one-tailed tests have more power to detect an effect in the specified direction but won’t detect effects in the opposite direction, while two-tailed tests are more conservative and appropriate when differences in either direction are of interest or when you want to guard against unexpected outcomes.