Robert Josephs of the University of Austin said that recently a research had revealed just how powerful and pervasive the influence of the endocrine system is on human behavior.

Results showed that individuals with elevated levels of testosterone and cortisol were more likely to overstate the number of correctly solved problems.

Having a little extra testosterone may give us the courage we need to cheat—and a little extra cortisol may give us the motivation.

Researchers asked people to take a math test and grade it themselves. The more problems they got correct, the more money they would earn. Saliva samples before and after the test showed that people with elevated levels of two hormones—testosterone and cortisol—were more likely to overstate the number of correctly solved problems.

“Elevated testosterone decreases the fear of punishment while increasing sensitivity to reward. Elevated cortisol is linked to an uncomfortable state of chronic stress that can be extremely debilitating,” says Robert Josephs, a psychology professor at the University of Texas at Austin.

“Testosterone furnishes the courage to cheat, and elevated cortisol provides a reason to cheat.”

Additionally, participants who cheated showed lowered levels of cortisol and reported reductions in emotional distress after the test, as if cheating provided some sort of stress relief.

Because neither hormone without the other predicted unethical behavior, lowering levels of either hormone may prevent unethical episodes. Prior research shows that tasks that reward groups rather than individuals can eliminate the influence of testosterone on performance; and, many stress relieving techniques such as yoga, meditation, and exercise reduce levels of cortisol, Josephs says.

“The take-home message from our studies is that appeals based on ethics and morality—the carrot approach—and those based on threats of punishment—the stick approach — may not be effective in preventing cheating,” Josephs said.

Researchers from Harvard collaborated on the study published in the Journal of Experimental Psychology.