CHAPTER 11 SOLUTIONS AND MINI-PROJECT NOTES
RELATIONSHIPS CAN BE DECEIVING
Both firewood sales and cough drop sales are likely to increase during the cold season.
No, it does not
imply that fires cause coughs!
No, you wouldn't necessarily conclude that walking causes lower weight.
People who prefer to exercise
(walk) may also be people who keep their weight under control, eat less fat, and so on.
could be possible but isn't likely, that milk-drinking alone is responsible for low risk of stroke.
is not likely, since it would say that a low risk of stroke causes people to drink milk, and vice
versa. People don't usually even know whether or not they are at risk for strokes.
possible, that milk-drinking is one of many factors that reduces the risk of a stroke.
likely, that confounding variables exist. One example is general good diet and good health, which leads to
reduced stroke risk.
is likely, that both reduced stroke risk and increased milk-consumption
result from a common cause, such as interest in taking care of one's health. Related to reasons 4 and 5,
note that the article made a point of saying that the same individuals who consumed the most milk were
also "leanest and most physically active."
, that both variables are changing over time, is not
relevant, unless the study did not adjust for age and the men drank less milk as they aged, which would
also increase stroke risk.
, that the association is just a coincidence, is possible. Any single study
could find a relationship just by chance.
See Figure 11.1 below.
If X-rays were deleted, the correlation would decrease (to 0.13 in case you are interested), because
outliers that fall in line with the rest of the data tend to increase correlation.
If pesticides were deleted (but
not X-rays) correlation would increase (to 0.68), because outliers that do not fit the pattern of the rest of
the data tend to decrease correlation.
That would decrease the correlation because it would be an extreme outlier, not in line with the data.
It also makes sense that it should decrease the correlation because it is an example of a case in which the
students and experts are not in agreement at all, and thus weakens the relationship.
Figure 11.1 for Exercise 11.4a
Winning time in an Olympic running event and the cost of running shoes are likely to be negatively
Page 1 of 4