How does one decide whether to use a one-tailed or a two-tailed test?
First, review the discussion of hypothesis testing. Our decision to use one-tailed or two-tailed tests depends on the nature of our null hypothesis, H0. If H0 maintains that some numerical characteristic of in the sample universe (i.e, a population parameter) is different from some constant (i.e., HA maintains the characteristic equals the constant), we typically use a two-tailed test. On the other hand, if H0 maintains the characteristic is less (greater) than or equal to a constant (i.e., HA maintains it is greater (less) than the constant), we typically use a one-tailed test.
Related Questions
- How do scientists decide which substances to test in animals, human laboratory cells, or human population studies?
- Can PROFILER help me test Nested Loops in my NATURAL programs? Can PROFILER help test IF/DECIDE Conditions?
- Where do students take the test? How does the College Board decide on test center locations?
