What happens when we get “hangry?”

On By Beth Morling, University of DelawareIn Chapter 10; Introduction to Simple Experiments, Correlational Studies, Experiments, Is this causal claim justified?

Screen Shot 2014-05-05 at 12.27.13 PMNational Public Radio covered a series of studies conducted by Brad Bushman and colleagues on an emotional state with a new name: "hangry."  It's when you are hungry, and more likely to get angry. Here's how Bushman describes the phenomenon:

"We need glucose for self-control," said Bushman, lead author of the study, which was released Monday in the Proceedings of the National Academy of Sciences. "Anger is the emotion that most people have difficulty controlling."  ..,Bushman said there's a good physical reason to link eating to emotion: The brain, which is only 2 percent of the body weight, consumes 20 percent of our calories.

According to the theory, when the brain is depleted of glucose, we have more trouble controlling difficult emotions such as anger.

In the study, about 200 married couples were followed daily for three weeks. Each night, their glucose levels were recorded, and they were given the opportunity to indicate their anger toward their spouse at that moment. How did they operationalize anger? By sticking up to 51 pins into a voodoo doll representing their spouse:

The researchers found that the lower the blood sugar levels, the more pins were pushed into the doll.  In fact, people with the lowest scores pushed in twice as many pins as those with the highest blood sugar levels, the researchers said.

a) Is the study described above correlational or experimental?

b)  Sketch a graph of the results they describe.

c) The headline of the NPR piece makes a causal claim–Study: Snack Might Help Avoid Fight With Spouse. Is the design of the study able to support a causal claim?

d)  Read the following comment:

Chris Beedie, who teaches psychology at the Aberystwyth University in the United Kingdom, said he thought the study's method was flawed ….The better way to test Bushman's concept is to give people high glucose on some occasions and low glucose on others, and see if that makes a difference in actual acts of aggression, he said.

What kind of study is Dr. Beedie describing? Why is it better than the real study's design? What would be the independent and dependent variables in Dr. Beedie's proposed study?

You can read the PNAS article here.