The culinary herb cilantro is very polarizing: Some people love it, but others hate it. A large survey of American adults of European ancestry found that 26% disliked the taste of cilantro.Suppose that this holds exactly true for all American adults of European ancestry. Choosing such an individual at random then has probability 0.26 of getting one who dislikes cilantro.

Use technology to simulate choosing many people at random.

Q1. Simulate drawing 25 people, then 100 people, then 400 people. What proportion, in each simulated sample, dislike cilantro? We expect (but because of chance variation we can't be sure) that the proportion will be closer to 0.26 in larger samples. Does this hold true in your samples?

Hint: the sampling applet or a Bernoulli Distribution can be used for this.

Q2. Simulate drawing a sample of 25 people 10 times (click "Reset" after each simulation) and record the percents in each simulated sample who dislike cilantro. Then simulate drawing a sample of 400 people 10 times and again record the 10 percents.

Provide an appropriate graphical summary of your results.

Which set of 10 results is less variable (the set of samples of 25 or the set of samples of 400)? Be sure to explain how you are measuring variability and describe how sample size is related to variability.