Answer:
B. A random sample of 2,800 U.S. adults under 35.
Step-by-step explanation:
To have an unbiased and accurate sample, the sample should be representative of the population about which we are inferring.
Also, the bigger the sample size, the less margin of error we will have in the inference.
As the population in this survey is about the US adults under 35, the sample should be selected from this group.
From the options that select from this population, the largest sample is 2,800.
So the answer is Option B.