Researchers wanted to determine whether pesticide Z is effective at killing grasshoppers that typically invade cornfields. Two 10-acre cornfields located 1 mile apart were chosen. The cornfield to the west was sprayed only with water

Question

The cornfield on the east received the same amount of water with a 10% solution of pesticide Z. The number of grasshoppers in the fields was recorded every week for 10 weeks. The number of grasshoppers in the west field treated only with water doubled after 10 weeks. The number of grasshoppers in the east field treated with pesticide Z declined by 50% after 10 weeks. Discuss two problems with the design of this experiment, and propose solutions to these problems.

Answer 1

First, the cornfields need to be closer together. Two cornfields a mile apart are more likely to have other factors such as soil, slope, and even precipitation differences that might affect the study. It would be more appropriate to study cornfields right next to each other. A second problem is with the sample size. You cannot perform statistical analyses with sample sizes of only one plot, so the conclusions drawn that are based on the results of the experiment would not be scientifically sound. The researchers would be better off subdividing the fields and replicating the treatment on several plots. They could then run statistical analyses and draw more solid conclusions as to whether the difference between treatment and control is meaningful. (An alternative potential problem a student might point out is the difficulty of counting every grasshopper in a 10-acre cornfield, with the repeated disturbance potentially damaging the crop, and with the unreasonable amount of time and person-power required. The student could then propose instituting some kind of representative sampling method, such as a series of small traps that are open for a short length of time.)