In: Statistics and Probability
1. Catherine’s the new CEO of an exercise equipment company and she wants to know if her time as CEO has affected the return rates of treadmills. Before she became CEO, 8% of treadmills were returned. What’s the null and alternative hypotheses of any tests she runs?
2. Imagine you were testing if a new battery lasts longer than the industry standard. You perform the appropriate hypothesis test with an unbiased sample and found statistical signifance at 99.9% confidence. It would be a mistake to claim that you “proved” this new batter lasts longer. Why?
(1)
H0: Null Hypothesis: P = 0.08 ( Her time as CEO has not affected the return rates of treadmills.)
HA:Alternative Hypothesis: P 0.08 ( Her time as CEO has affected the return rates of treadmills.) (Claim)
(2)
H0:Null Hypothesis: ( a new battery does not last longer than the industry standard. )
HA: Alternative Hypothesis: ( a new battery lasts longer than the industry standard. ) (Claim)
= 0.001
Since the difference is significant, our conclusion is :
There is sufficient evidence from the available test data to support the claim that a new battery lasts longer than the industry standard..
By this hypothesis test, we do not conclude that the Alternative Hypothesis: a new battery lasts longer than the industry standard is proved.
We are just concluding that the available data provides statistical evidence against the null hypothesis: a new battery does not last longer than the industry standard. This is all.
Further technical analysis is required to prove this new batter lasts longer..