This paper sounds dubious. Tiny sample size. The subjects were all women. Trust is very subjective. The language they use indicstes a lot of bias. If testosterone has any effect it will be tiny compared to other factors like ethnic and religious homogeneity, personal integrity, life experiences, personal wealth or the ability to absorb losses. Any chemical change will cause stress and that will lead to anti social effects. This kind of test, if done properly, would need to control for thousands of effects. I can see how there would be an urge to simply accept it becuse it confirms some experience we have had with roid raging morons, or aggressive high test niggers but honestly those subjective experiences are as scientific as this paper.