A scientist is designing a study to compare clover accumulation for ve dierent planting densities. A eld

is divided into 20 plots with 4 plots assigned to each density. Why is it necessary to randomly assign the

planting densities to the plots?

A. To maximize the potential of determining dierences.

B. To minimize the heterogeneity in the variances.

C. Randomization is necessary since all the plots are in the same eld.

D. To avoid government intervention.

E. To provide appropriate reference populations for statistical inference.