If this question is on the final exam, I will provide you with a copy of figures 7-4 and 7-5 of the methylmercury risk assessment.

1. The data graphed in Figure 7-5 of the Methylmercury risk assessment show huge scatter. What errors in gathering and analyzing the data are responsible for this scatter? (one paragraph answer)

Answer: The scatter is not primarily a consequence of errors made in gathering or analyzing the data. Rather, it is a consequence of the fact that the correlation between methylmercury exposure and cognitive ability is very weak. Even if the data were gathered with high precision, high accuracy, and with the utmost care, the graph would look basically the same as that shown. (I admit that this is a bit of a trick question. But now that you have the answer, it is certainly fair game for an exam.)

2. The data graphed in Figure 7-5 of the Methylmercury risk assessment show huge scatter. What is the primary cause of this scatter?

Question note: This question overlaps the previous question, though not entirely.

Answer notes: It would be incorrect to cite error in collecting and gathering the data as the source of this scatter. The Figure shows that, evidently many factors other than methylmercury exposure affect cognitive ability. There is little information in the methylmercury risk assessment as to what these other factors might be, though parental education level, parental score on cognitive tests, prenatal exposure to ethly alcohol, and educational opportunities are all obvious possibilities.

3. Figure 7-4 shows that a graph of exposure vs. dose might show one of several theoretically possible forms. Do the data in Figure 7-5 convincingly refute any of the theoretically possible models shown in Figure 7-4? (one paragraph answer)

Answer hint: The data graphed in Figure 7-5 are consistent with many different models, including several of the theoretical models shown in Figure 7-4. One can see this by seeing that the authors have drawn several lines through the data in Figure 7-5---these data evidently fit several models equally well. To finish answering the question, you will have to determine if any of the models shown in Figure 7-4 can be excluded.