A large company event was held in a park on a hot day. Afterward, many of the employees got sick. Some of the employees blamed the potato salad. To investigate, a random sample of 20 sick employees and a random sample of 20 non-sick employees are selected. Each selected individual is asked if they ate the potato salad. The results are displayed in the tables.

Observed counts:



Expected counts:



Management would like to know if there is convincing evidence that the distribution of response about eating the potato salad differs for all employees who are sick versus those who are not sick. They decide to test these hypotheses:

H0: There is no difference in the distribution of responses among those who are and are not sick.
Ha: There is a difference in the distribution of responses among those who are and are not sick.

The conditions for inference are met. The chi-square test statistic is χ‑2 = 12.13 and the P-value is less than 0.0005. Because the

P-value is small, there is convincing evidence of a difference in the distribution of responses among those who are and are not sick.

For which cells do we observe more employees than expected?

More sick people than expected say that they ate the potato salad, and more non-sick people than expected say that they ate the potato salad.
More sick people than expected say that they ate the potato salad, and more non-sick people than expected say that they did not eat the potato salad.
More sick people than expected say that they did not eat the potato salad, and more non-sick people than expected say that they ate the potato salad.
More sick people than expected say that they did not eat the potato salad, and more non-sick people than expected say that they did not eat the potato salad.