[MUSIC PLAYING]

NARRATOR: Scientists and researchers have many ways of answering questions. Observation, case studies, and surveys can help gather information and isolate trends. But those methods can't tell us if one variable causes another. To find cause and effect, scientists and psychologists conduct experiments, procedures based on the scientific method undertaken to support or contradict a hypothesis or a testable prediction.

RANKIN MCGUGIN: The hypothesis is your idea of what's going to happen, what you think you're going to find when you perform an experiment. So your initial hypothesis, why you begin the whole experiment, really guides you throughout the whole procedure.

And here is a screening form that we need all of our MRI participants to fill out.

NARRATOR: Dr. McGugin is overseeing an experiment that explores the fusiform gyrus, a pea-sized region in the brain thought to be responsible for facial recognition.

RANKIN MCGUGIN: There's been a big controversy in whether this region is exclusively responsible for face perception or whether it can actually be modeled and changed through development based on our experience with different object categories. OK,

So in this behavioral part of the experiment you're going to test your knowledge of cars.

NARRATOR: To test the hypothesis that the fusiform gyrus can develop expertise beyond facial recognition, Dr. McGugin is examining the brain activation patterns in people who are experts in a field, in this case, car buffs, versus non-experts.

RANKIN MCGUGIN: So you'll press one, two, or three.

NARRATOR: In a traditional psychology experiment, testing a hypothesis involves manipulating one or more factors called the independent variables and observing the effect on the dependent variable.

RANKIN MCGUGIN: The independent variable is the variable that you set. So it's the one that the experimenter controls as opposed to the dependent variable. So car expertise, the variants in car expertise here would be our independent variable, and the dependent variable would be the brain activation in the fusiform gyrus when subjects look at cars.

So I'm going to leave now. If you have any questions, you can come get me.

NARRATOR: After coming up with a hypothesis and defining the independent and dependent variables, scientists recruit participants and assign them to experimental and control groups.

RANKIN MCGUGIN: The experimental group is a group that varies in the factor that you want to measure. For our experiment we want a group that varies in car expertise. The control group really serves as a baseline. In our control, for example, would being a group of subjects that did not vary in car expertise, and we would expect to see a difference between the experimental group and the control group.

NARRATOR: Although not employed here, many psychology experiments randomly assign subjects to the experimental and control groups to minimize the possibility that an unknown variable is confounding the results.

RANKIN MCGUGIN: After the behavioral tests, which are done in the lab, we then move to the MRI machine. It's then up to us to analyze the data. So we would expect that subjects who do really well on the behavioral car task could also have very high levels of activation in the fusiform base area in response to cars. And that's what we find over and over and over again, and that this area can be changed with learning.

The black and white image is the anatomical and structural image, and then overlaid on top of that—

NARRATOR: In this experiment, the data appear to support the hypothesis. But that's often not the case.

RANKIN MCGUGIN: The measure of a good experiment is definitely not finding what you thought you were going to find. That happens sometimes, but oftentimes the most interesting results come when you least expect them. So it's very important not to go into an experiment with set expectations and try to find those certain things because you have to be honest and open about the results.