A common type of expertimental study involves the comparison of an experimental sample and a control sample ; for example, an experimenter interested in the effect of a plant hormone on the growth of beanstalks might treat alternative rows of beans with the harmone, leaving half the plants untreated. The reasher will measure the height of each beanstock and then use a statistical test to determine if the difference in mean height between the two samples of beanstalks is significant. Selecting one bean for the treated group does not give any information about what beans will be chosen for the other group. Would selecting one dieter's 'before" weight tell anything about the composition of the "after" sample?
In which problem above would you say the two samples are independent?
(a) The beam problems
(b) The dieting problem