Question

Amazon is conducting an A/B test for its shopping cart to analyze Total Units Ordered (TUO), representing the number of items ordered. State the null hypothesis (H0) and alternative hypothesis (H1) for this A/B test.

a) H0: There is no difference in Total Units Ordered between the shopping cart versions; H1: There is a significant difference.

b) H0: The new shopping cart increases Total Units Ordered; H1: The new shopping cart has no effect.

c) H0: Total Units Ordered is higher in the new shopping cart; H1: Total Units Ordered is the same.

d) H0: The old shopping cart outperforms in Total Units Ordered; H1: The new shopping cart is better.

Answer 1

In the A/B test for Amazon's shopping cart, the null hypothesis (H0) posits no difference in Total Units Ordered between different versions, while the alternative hypothesis (H1) suggests there is a significant difference.

Step-by-step explanation:

When conducting an A/B test on Amazon's shopping cart, the effect on Total Units Ordered (TUO) is being analyzed. The correct formulation of the hypotheses should establish a baseline (null hypothesis) that assumes no effect and an alternative hypothesis suggesting there is an effect. Here, option a) presents the correctly stated hypotheses:

H0: There is no difference in Total Units Ordered between the shopping cart versions.

H1: There is a significant difference in Total Units Ordered between the shopping cart versions.

This is because Neyman-Pearson null hypothesis testing (NHT) starts with the assumption that there are no differences between the groups being tested (the null hypothesis). If we find evidence to the contrary, we can consider that there is a statistically significant difference, thus supporting the alternative hypothesis.