The Flawed Experiment That "Proved" Stress Is Good For You

We may earn a commission from links on this page.

The Yerkes-Dodson Law predicts how people react to stress. It's also one of the most contested — and skewed — laws in psychology. Find out about the 1908 experiment that shocked lab mice, and then the world. In that order.

The Yerkes-Dodson Experiment

I remember, back in health class, hearing that, while a too much stress destroyed a person's performance, a little bit of stress made them perform at their best. It made them alert and ready to spring into action. I didn't quite believe it then, and having gone into the fast-paced, high-stakes world of writing in my pajamas, I don't believe it now. It turns out that the question wasn't nearly as settled as the health teachers at my school made me believe.


The entire story goes back to 1908, when Robert Yerkes and John Dillingham Dodson decided to engage in that most popular activity of psychologists — being horrible to lab rats. These particular lab rats were made to choose between a black door and a white door under different lighting conditions. They were corrected, when they made the wrong choice, by a series of electrical shocks. Yerkes and Dodson noticed that mild shocks improved the rats' performance. Slightly increasing the shocks increased the performance, until a certain point, after which it greatly decreased the performance. A chart of the shock-strength versus performance forms an inverted U. The Inverted U shape became the standard for modeling performance under stress. A little stress, and performance improved. Too much and it suffered.


The Yerkes-Dodson Law

This idea of stress improving performance, up to a point, became a truism in the psychological community. It was called the Yerkes-Dodson Law. It's not Yerkes' or Dodson's fault that became so famous. In the original experiment, they never claimed to know the inner workings of the rats' minds, they simply charted how they performed. It was others who attributed their performance to stress, and then supposed rats and humans had similar responses.


Through the years, researchers have repeated the experiment, sometimes with other animals, and found all kinds of different responses, depending on species and the exact set-up of the experiment. Some animals grow accustomed to the shocks physically, and stop responding to them, but their overall performance suffers. In that case, increase in stressors does nothing to improve performance. It's just a steady slope downhill. At the same time, similar experiments were being done with rats showed traditional Inverted U outcomes. Rats that were oxygen-deprived followed the curve on tests. Rats given swim tests also fit the curve. Rats, overall, seem to be very consistent with an inverted U.


What about humans? Many organization have studied performance under stress, including NASA. (Wouldn't it be amazing if Buzz Aldrin got his name for doing well in a Yerkes-Dodson Test?) Overall, results show that humans are... complicated.

How we respond to stress depends on our goals, our emotional make-up, and the kind of stress involved. It's not surprising that, if the stress in question is sleep-deprivation, no amount of it will improve performance. There's also the question of what performance you want to affect. Studies found that loss of sleep will slow a person down, but surprisingly won't do too much to their accuracy. Heat and noise, on the other hand will take out accuracy, but not speed. Different stressors do different things to our ability. It's better to gather specific data than rely on a general law.


Top Image: Alexey Krasavin

[Via Stress, Cognition, and Human Performance, From the Inverted-U to the Extended-U.]