Key Study: Testosterone and aggression in rats (Albert et al. 1986)

Travis DixonBiological Psychology, Criminology

Castrating animals reduces their levels of aggression as it prevents the production of testosterone. This has been observed in numerous rat (and other animal) studies.

The material here was originally published in IB Psychology: A Student’s Guide.

Males are more aggressive than females in almost every species. Why is this? One possible reason could be in the influence of the male sex hormone, testosterone. Testosterone is produced in the testes of males, which could be why males are more aggressive than females. While it is also produced in the ovaries of females, this is at a much lower rate. Perhaps it doesn’t come as a surprise that there have also been correlations found between violent crime and testosterone levels (e.g. Dabbs et al., 1997 as cited in Batrinos, 2012; you can read more here).

Numerous animal studies have shown that castration (cutting out the testicles) will reduce the aggressive tendencies in males. The following study by Albert et al. (1986) is a good example of a well-controlled experiment that demonstrates a causational relationship between testosterone levels and aggression.

Word to the wise: this study doesn’t show how testosterone influences aggression, only that it does. Other studies in the criminology chapter (e.g. Goetz et al. and Radke et al.) provide better evidence explanations for the relationship between testosterone and aggression.

Albert et al. (1986) wanted to investigate the effects of changing testosterone levels on the aggressiveness of male rats. They placed rats in cages and identified the alpha males. An alpha male is the leader of the colony. In animals, this is typically the biggest and strongest. The term can be applied to any animal group, including humans. So the researchers identified the alpha males and they measured their aggression levels when there was a non-aggressive rat placed in the same cage. They measured aggression by recording behaviours such as biting.

After they measured the aggression levels they divided the alpha males into four
separate groups to undergo four separate surgeries:
A. Castration
B. Castration followed by implanting of tubes with testosterone
C. Castration followed by implanting of empty tubes
D. A “sham” castration followed by implanting of empty tubes (this means they would have cut open the rat and sewn it back up without actually removing the testicles).

They then measured the change in aggression when non-aggressive rats were introduced to the cage. Those that had the operations that reduced testosterone levels (e.g. Group A and C) had a decrease in aggressiveness (e.g. attacking and biting) but those that had the operations that kept testosterone levels in tact (Group B and D) didn’t have a significant change in aggression levels.

This evidence by itself demonstrates a correlation between testosterone and aggression. It was followed by a second operation so that those that had the surgery that decreased testosterone had another operation that increased testosterone (e.g. Group C had their tubes filled with testosterone). Those alpha rats that had their testosterone replaced showed returned levels of aggressiveness similar to those in the “sham” castration group.

Moreover, the researchers observed that when a subordinate male (one that is not the alpha) is placed in the same cage as an alpha rat that has been castrated the lower rat (subordinate) becomes the dominant (alpha) rat in the cage. Also, when a rat that had the sham operation is put in a cage with a castrated rat, the sham operation rat shows higher levels of aggression. This suggests that testosterone may facilitate behaviour associated with social dominance in rats. By experimenting on rats, researchers are able to determine correlations between biological factors and behaviour. Albert et al. were able to manipulate levels of testosterone and conclude that levels of testosterone affect aggression.

You can see the advantages of using a laboratory experiment to manipulate variables in Albert et al.’s study. By measuring effects of aggression on manipulations of testosterone levels, a clear cause-effect relationship can be determined.


Albert, D. J., M. L. Walsh, B. B. Gorzalka, et al. “Testosterone Removal
in Rats Results in a Decrease in Social Aggression and a Loss of Social
Dominance.” Physiology & Behavior. U.S. National Library of Medicine.

You can find the original study here.