The question as whether men are innately better at chess than women is controversial. At least, this idea is widespread, to the point where it has become a stereotype : men are better at chess than women.
An interesting article that stresses this point is the one by Anne Maass et al., who conducted a study in Italy. I will discuss this study further, but before entering the subject, here is the reference for those who are interested in reading the full article :
Maass, A., D’Ettole, C., & Cadinu, M. (2008). Checkmate? The role of gender stereotypes in the ultimate intellectual sport. European Journal of Social Psychology, 38, 231-245.
Women chess players represent less than 5% of all tournament players worldwide. Added to that underrepresentation of women is their underperformance: they are only 1% of all grandmasters. It is easy to say that men are more interested in chess for one, and better at it for two. The aforementioned study tends to go in the other direction.
The goal of Maass et al. was to investigate the role of the widespread stereotype (men are better than women at chess) in the women’s underperformance. It seems that it’s the first time researchers investigate another venue than cognitive differences to explain the better performance of men in chess.
First, the authors lead a pilot study in order to verify that the stereotype is indeed widespread. They found that the participants of the pilot study did think that men were innately better at chess than women. They also found that the participants thought other players too had the idea that men are superior to their female colleagues.
In the study, Maass et al. explicitely tested the impact of the activation (mention) of the stereotype before a game. They also checked for the importance of cognitive differences in the results obtained (for instance, it is well established that men score better in mental imagery and mental rotation, so they wanted to check if these cognitive differences could explain more the difference of results than the mention of the stereotype would).
The experiment. The study was conducted via Internet. Fourty-two male participants and 42 female participants were paired according to their Elo rating for the games, i.e. each woman was paired to an equal-strength man. There was a maximum difference of 30 points between the man and woman’s ratings. All in all, this means each player was of the same strength. These pairs of players played 2 games against each other, but without knowing they were playing twice against the same opponent. They just thought it was two games against two different opponents, because the nicknames weren’t the same in both games (they were neutral-gender nicknames, so that guessing the sex of the opponent would be impossible).
There were two conditions in the experiment. In the control condition, participants were just playing two games with their paired opponent of the other sex (still without thinking it was the same opponent twice). They didn’t know whether the opponent was a male or a female, and the experimentators never mentioned anything about men being superior to women. In the experimental condition, the participants were told in the first game that they were playing against a man, whereas in the second game they were told they were playing against a woman (that order was reversed for half of the participants). Before the game, the participants in the experimental condition were told "that recent studies had shown that men earn clearly superior scores than women in chess games. The aim of this study is to provide further tests for these findings". That was the way to activate the stereotype for the participants in the experimental condition (as I said, the participants in the control condition weren’t told anything about that stereotype).
So, what were the results of the games?
First of all, women in the control condition played just like anyone else (they scored approximately 1.0/2.0). Women in the experimental condition also played as expected when they thought they were playing against another woman (1.0/2.0), but their performance dropped drastically when they thought they were playing against a man (they scored around 0.5/2.0) (see figure below, adapted from Maass et al., 2008).
- 
Moreover, the authors checked for the participants' intention of playing aggressively for the win before their game. In the control condition, there was no difference between men and women as to how much they planned on playing for the win (as opposed to playing to avoid making mistakes). There was no sex difference either when the women in the experimental condition were told that they were playing against another woman. But when the women in the experimental condition were told that they were playing against a man, they planned on playing much more defensively, and effectively ended up doing so. The participants’ results in their games was positively correlated with their intent to play aggressively, so this could partly explain the women’s underperformance in chess : due to the activation of the stereotype, they play overly defensively.
Also, results suggest that men are not cognitively at advantage over women, since there was no significant correlation between mental rotation or imagery capacity, and results in the games.
Moreover, despite the women knowing they had the same Elo rating as their opponent, they showed a lower chess-specific self-esteem, which too was positively correlated with the results in the games. This lower self-esteem is most likely at least partly due to the knowledge of the stereotype : this could also explain the underperformance of women in chess.
It is to note that women score higher on individual tests evaluating chessic abilities than they do in tournaments, whereas it is NOT the case for men (in their case, both scores match). It is very plausible to believe, given all this data, that women would score just as much as men in chess tournaments would it not be for the existence of the pervasive – and false – stereotype that leads people to believe (women included) that men are better than women at chess.
