Conducting an experiment requires you to do something to members or units
of a population and then to record their response to the treatment that
you have imposed. Unlike observational studies, in experiments you are
searching for evidence that the treatment caused a specific response.
If you are examining two variables, it is likely that you will want to
make a claim about the effect of one variable on another. In other words,
one variable caused, at least to some degree, a change in another variable.
Which way does the causation go?
Let's look at an example. We will examine whether babies fed on breast
milk are more or less likely to be ill.
Which one of the situations below seems more likely to you, A or B?
A. Feeding a baby on breast milk causes resistance to
disease.
B. Resistance to disease causes a baby to feed on breast
milk.
Which variable – 'feeding a baby on breast milk' or 'resistance
to disease' – caused the other? This is the sort of question that
is asked in experiments. Most people would say that feeding the baby on
breast milk caused the baby to become resistant to disease.
A variable that can be used to explain, or can be said
to cause differences in another variable is
called an explanatory variable. The variable in which
the differences are observed is called the response variable.
