Answer
The scientist should put together a retrospective study in which she analyzes medical data of heart attacks in a population of people. Preferably, the population she selects should be fairly uniform (e.g., adult men, between the ages of 25- 50, within a certain weight range, with no pre-existing heart issues) in order to limit the number of variables that could have led to the heart attack. These common variables amongst the population are the control.
From that population of men with heart attacks, she could then analyze how many of them were using the drug which she believes leads to heart attacks. She would then perform statistical analysis to determine if the numbers of heart attacks are statistically higher in the men taking the drug in question.
The experimental variable is the common drug. The heart attack suffered is the responding variable.
Work Step by Step
It is obviously not practical to conduct a study with a drug that could potentially induce heart attacks in study participants. As such, it is best to base a study upon a statistical analysis of a population of heart attack victims. Within that population, selecting a number of controlled variables such as sex, age, weight, etc. would help to limit the number of things that could have influenced the cause of the heart attack. Such limitation helps increase the statistical significance if a heart attack occurs in people who are taking the drug in question vs those who are not.
