a. No. People cannot be randomly assigned to attend cultural events or not.
b. We cannot conclude the relationship is causal because this is an obervational study and there may be other confounding variables. There could be other factors that account for the increased longevity, for instance, maybe those who are already healthier are more likely to attend cultural events.
c. The article implies that there is a causal connection and seeks as explanation; the causal link is not justified.
d. General health at the start of the study and amount of exercise in their daily lives. Maybe those who are healthier and who have more energy are more likely to exercise and to attend cultural events.
Problem 6.
It is easier to detect a difference between two groups if there is little natural variability amoung the measurements within each group. By dividing the babies into low, medium, and high birth weight groups, Salk was lowering the natural variability within the groups.
Problem 11.
The board assumed that when students answered a question wrong each of the choices were equally likely to be selected (random guessing). This is generally not true in multiple-choice exams, in which certain answers can be ruled out and others are plausible, although incorrect.
Problem 18.
a. It would be easier of they were all the same. Otherwise it might be hard to distuinguish the difference due to grades from natural variablility introduced by using different car sizes. For instance, suppose one grade gave 2 more miles per gallon (mpg) than another. If all of the cars got around 33 mpg with one grade, and around 35 mpg with the other, the differences would be noticed. If instead they ranged from 15 to 40 with one grade and 17 t 42 with the other, it would be harder to notice the difference.
b. One disadvantage would be extending the results to cars of other sizes. There would be no way to know if they could be extended.