So that's the underlying concept of why we do an experiment, now we just want to add on a little of the vocabulary that scientists use. So when I'm talking about experiments, they call the variables in the experiments by special names. One of them they call the independent variable, and this is the variable that is being directly manipulated by the experimenter. So we're holding everything constant, no variables are changing, and then we pick one thing and we change it to see if it causes a difference in the other variables.
So in our very ridiculous example, the independent variable would be ice cream, because we're holding everything else constant, and then we're changing whether we're giving people ice cream or not to see if that causes any kind of an effect.
The other variable is called a dependent variable, and that's the variable that's being measured by the experimenter. Now, this is a little hard to remember in terms of what's independent variable, what's dependent variable. The way that I remember it is by looking at the term dependent variable, because that one actually makes some sense. It's called the dependent variable because we are wondering if it depends on the independent variable.
So the dependent variable in our stupid example would be number of snake bites. In other words, does getting bitten by a snake depend on whether you've eaten ice cream? It's just another way of asking, does eating ice cream cause you to get bitten by a snake? But for some reason, we choose to flip it around in the terminology. If we wanted to use the same terminology, we might call the independent variable, the variable that we think might be causing changes in other variables, and we'd call the dependent variable, the thing we think might it be getting changed by the thing that we think might be causing changes, right? But that's a very awkward way to talk about it, so we just call them independent variable and dependent variable.
Again, the idea is the independent variable is kind of independent of everything. Nothing's causing it to change, except when we go in and directly manipulate it. But then the dependent variable is the one that we think might be dependent on the other variable. It might be being caused to change by the other variable.
So if this were the study of vegetarianism, then the independent variable would be diet type, right? And the dependent variable would be your health. So we're going to directly manipulate the kind of diet that you have, and then see what your health outcome is. So this would be a way of going beyond the correlational study that we looked like and turning it into an experiment. But, of course, it's much harder to do on a widespread scale and for a long period of time. To do this kind of experiment, you'd have to get a whole bunch of people to agree to radically change their diet, which is very difficult to do, and you'd have to randomly grab people—half of the people who agreed to do this, and have them eat one diet, have the other people eat a different diet, and then you'd have to follow them up for potentially 10, 20, 30, 40, 50 years to see what the long-term impact is on their health.
So we do conduct experiments like this, but they're very expensive, but we do them because they can yield much more valuable results. We can actually say something about whether this particular diet causes a difference in your health outcome, whether this independent variable causes the difference in the dependent variable.