Answer:
Null hypothesis: H0:μcontrolOPS=μtreatmentOPS
Alternative hypothesis:
Ha: μcontrolOPS≠μtreatmentOPS.
Step-by-step explanation:
1)The Null Hypothesis: The null hypothesis (H0) posits that the average OPS of the control group aligns with that of the treatment group. This can be formulated as:
H0:μcontrolOPS=μtreatmentOPS
2)Alternative hypothesis: The alternative hypothesis (Ha) indicates that the average OPS for the control group differs from that for the treatment group, meaning the average for the treatment group is not equivalent. It can be expressed as Ha: μcontrolOPS≠μtreatmentOPS.
This is a two-tailed hypothesis test, examining whether there is a variance in either direction. Amazon's inquiry is whether the average of the control group is either higher or lower compared to the treatment group. Utilizing a one-sided test would be inappropriate in this case. A one-sided test would only be applied if Amazon believed the treatment might yield the same or a better average OPS than the control.
If using a one-sided hypothesis, it would suggest that the average OPS for the control group is at least as high as the average for the treatment group, expressed as:
H0: μcontrol OPS≥μtreatment OPS. The alternative would then be that the average OPS is less for the control group, indicated as:
Ha: μcontrol OPS<μtreatment OPS. Mixing a null hypothesis stating equality in both groups (H0: μcontrol OPS=μtreatment OPS) with an alternative hypothesis asserting that the average OPS of the control group is less than that of the treatment group presents a blend of one-sided and two-sided tests, which is incorrect.