the one that supports the use of large, representative samples in a study is :
B. the traits of the sample members may differ randomly from the traits of the population as a whole
Which means that, taking only small number of samples wouldn't be enough to represent the accurate data to find out the cause and effect
hope this helps