Men who were administered testosterone lied less, even when no one was watching.

PROBLEM: The association between testosterone and aggression isn't as simple as many are lead to believe. Game theory experiments have suggested that testosterone actually causes people to be more "prosocial" (meaning that they voluntarily act in the interest of helping others), but the exact relationship between the hormone and behavior is unclear. 

Does testosterone directly cause people to act in the interest of others? Or does it just  make them more concerned about their social status, thus making them appear to be acting in the interest of others while really working to improve their image? Alternatively, does testosterone simply affect their beliefs about other players in the game, which then causes them to alter their behavior?

---

• Republican Congresswomen Look Twice as 'Feminine' as Democrats
• Castration Adds Years to Men's Lives
• Oxytocin ('The Hormone of Love') Also Makes Us Conformists

---

METHODOLOGY: Forty-six men had a testosterone-containing gel applied to their skin, and 45 others received a placebo. The next day, after the hormone had been given time to sink in, they participated in an experiment where they repeatedly rolled a die and were paid according to how high they scored.

The results were self-reported, so the subjects were free to be as honest or dishonest as they liked about their outcomes. In order to completely eliminate their fear of getting caught (or, conversely, their desire to appear honest), the study was designed so that it was actually impossible for the researchers to find out whether the subjects were lying or not. Instead, they used the power of statistics to see if anyone reported more earnings than was probable, and inferred from there how honest the men were being.

RESULTS: Both groups' earnings skewed right, meaning that both engaged in self-serving lying. However, the testosterone group reported significantly lower payoffs than the placebo group, implying that they were more honest (or, less dishonest). In addition, blood tests conducted after the experiment indicated that overall testosterone levels were associated with lower reported payoffs.

CONCLUSION: The results suggest that there is a causal relationship between administration of the hormone and increased truth-telling. It's basically the muggle version of veritaserum.

IMPLICATIONS: These findings support the theory that testosterone is associated with pro-social behavior in men. Because the study controlled for social influence, it supports the first (and most optimistic) hypothesis for why this is so: testosterone appears to genuinely decrease self-interest.

The full study, "Testosterone Administration Reduces Lying in Men" is published in the journal PLoS ONE.