Given that A
dataset shows that, on average, smokers do not live as long as
nonsmokers and heavy smokers do not live as long as light smokers.
If
you compute the least squares line for y = age at time of death and x =
number of packets per day typically smoked, you will notice that as the number of packets per day a person smokes increases, the age at time of death of the person decreases.
Thus, the graph wil have a negative slope.
Therefore, the slope of your
regression line will be less than 0.
