Question

a) How can we decide if we need to use a one-tailed or two-tailed test when conducting our analyses?

b) What are the differences between the two?

c) Why do we need to make the choice between a one-tailed or a two-tailed test? Discuss types of research where using the t statistic may be an appropriate alternative to using a z-score

Answer 1

a. If the alternative(Research) hypothesis contains words like greater than, less than , Then test will be one-tailed test. e.g, Ha: Proportion<0.7, Ha: Mean >9.2 etc.

If the alternative(Research) hypothesis contains words like different form or there is difference, Then test will be two-tailed test. e.g, Ha: Proportion≠0.7, Ha: Mean ≠9.2 etc.

Hence,

We decide test is one-tailed or two-tailed from the research hypothesis(Ha) only.

b. The basic difference between one-tailed and two-tailed test is that, the symbol like "<", ">" in research hypothesis of one-tailed test and symbol like "≠" in two-tailed test.

c. To find critical value for the test, we have to make the choice between a one-tailed or a two-tailed test.

We can use t statistic over z-score where the research contains Non-normal population,Small Sample size and Unknown population parameter.

Thank you.