Stephanie believes that the pills helped her to learn and remember her chemistry, but we don't know what grade she would have gotten on the exam without the pills. In order to decide if the pills improved memory, we need a double blind study in which many students (selected by flipping a coin or some other random method) receive the memory pills (this is the experimental group) and other students receive pills that look like the memory pills, but really are "sugar pills" or placebos. This is the control group. The participants should not know which pills they are getting so their beliefs don't affect the results. The experimenters should also not know who is getting the memory pills or the placebo so that their expectations don't affect the results either. Every participant in both of the groups would then take the same memory test. What results would you expect if the memory pills really worked? What if they didn't?

Of course, people are different in how much they remember, so we expect that some will score high on the memory test and some low. The evidence for or against the memory pills comes in comparing the average scores between the two groups.

You would need an answer to the following questions in order to decide if the memory pills "worked."

1. Did the group taking the memory pills (experimental group) remember more than the group taking the placebo (control group) on average? (In other words, don't compare one score from each group, average out all the scores from each group and then compare them with all the averaged scores from the other group.)

2. Was the difference in the scores between the two groups large enough to be important? ("Large enough" has a precise meaning in statistics that concerns different types of significance.)

3. Have the results been replicated? When results are replicated, the same results are obtained when the study is repeated with new participants.