A) The number of eight-year-old children in the United States was not substantially smaller for the year 2000 than it was for the year 2014.
It doesn't matter if the sample pool changed because we have a proportion in our hand and even if the sample size changes it shouldn't matter statistically
B) Eight-year-old children were the only group of children diagnosed for autism between 2000 and 2014.
Argument and the numbers revolves around 8 years old children. Even if they diagnosed children from other age groups it wouldn't effect our numbers and argument because numbers "one out of every 150 eight-year-old children" & "one out of every 68 eight-year-old children" would still not change as they only consider 8 years old
C) The most effective age at which to test children for the incidence of autism is the age of eight.
Still it wouldn't make any difference because we are comparing children in same age group.
D) There were no important changes in the criteria for diagnosing autism in eight-year-olds between 2000 and 2014.
CorrectFor example if you applied in 2000 criteria x, y & z but applied in 2014 criteria x & y to determine autism in children, you may have significant changes in your numbers than what you would normally have with criteria x, y & z.
E) Autism is not the only condition for which per-capita diagnoses increased substantially between 2000 and 2014.
Irrelevant, we don't care about other conditions