A common, but erroneous idea regarding testosterone is that it works like dry wood on a campfire: add more wood and the fire burns higher and hotter. It doesn't seem to work that way at all (Bjorkqvist, 1994)....[C]ertain behaviors and experiences seem to affect testosterone levels, rather than the other way around (Brody, 1999; Susman, Worrall, Murowchick, Frobose, & Schwab, 1996). This seems to be a difficult reality for many to grasp. Moreover, individual fluctuations in testosterone levels are not related to fluctuations in aggressive behavior. In other words, a person is
not more likely to behave aggressively when testosterone level is higher. In situations of social stress (such as fights for status), testosterone levels may rise, but so do the levels of several other hormones (cortisol, for instance). To further complicate things, you need to know that testosterone is usually transformed...into other substances, such as estrogen or estradiol, for example. Higher levels of estrogens are also correlated with elevations of aggression (Brody, 1999).
...[E]xperience and resources seem to modulate any effects of testosterone. Continuing this line of research, Mazur, Booth, and Dabbs (1992) studied the relationship between testosterone levels and the outcomes in a chess competition. In postgame measurements, winners had higher testosterone levels than did losers. However, the eventual
losers in the finals tournament had significantly higher testosterone levels
before the matches.
...Overall, testosterone levels seem related to strength, impulsiveness, and adventurousness and these can be shaped into positive or negative behavioral characteristics according to the prevailing gender ideology.