Question

I have the task of comparing paired samples of a customer waiting time from an old process versus a new process. The manager would like the test set up to see if the wait times using the new process is shorter. I have to use the Six Step Process,

YOUR EXPERT ANSWER: When comparing the customer waiting time from an old process vs a new process, it is appropriate to use two tailed hypothesis in this case. Null is wait time 1 = wait time 2 and Alternate is wait time 1 media%2Fdde%2Fddec02e6-0a21-4744-93d1-63 wait time 2 so, two tailed hypothesis is better option when comparing two positions using t distribution test. two tailed hypothesis test will check for lower tailed test as well as upper tailed test.

My second question: Even though the manager wants the test set up to see if the wait time using the new process is shorter, will the null and alternate still be equals to and not equals to?

Answer 1

When the manager wants the test set up to see if the wait time using the new process is shorter, the appropriate hypotheses are

Null Hypothesis H0: Average wait time of new process is equal to Average wait time of old process

Alternate Hypothesis H0: Average wait time of new process is less than Average wait time of old process

Please note that whatever we are claiming or testing should reflect in alternate hypothesis. Null hypothesis mostly contains equal to operator and assuming that null hypothesis is true, we try to estimate the probability of finding the observed, or more extreme, results (p-value) to make a conclusion on alternate hypothesis.