Gender and competition in adolescence: task matters

Based on the findings, girls were found to be more altruistic, less risk taking, and equally as competitive as boys, except in the area of math.

• Boys performed better in the math task than girls did, but there were no performance differences by gender on the verbal task.  
• While boys were 19% more likely to choose to compete in math than girls, when performance beliefs were controlled for, this difference became insignificant. There was no difference in choice to compete in verbal exercises between genders.
• Relative to their true performance, girls were, on average, underconfident about their performance in both the verbal and math areas. Boys were slightly underconfident in math, but not in the verbal task.
• Girls were more risk-averse than boys: boys were more likely to participate in the all-or nothing lottery than girls. Boys also self-reported higher tendencies to take risks.
• Girls were more altruistic than boys, giving SEK 6 ($1 US) more to charity on average.

In short, it is suggested that the gender gaps in risk taking and altruism- two preference differences offered as contributing factors to wage and labor market gender gaps- begin in childhood or adolescence. The study also suggests that there is no difference in competiveness between genders, except in math, which can be attributed to preconceptions regarding personal performance. Thus, no definite conclusion can be drawn regarding competitiveness.

Researchers chose nine classrooms in five high schools in the Swedish cities of Stockholm, Uppsala, and Vasteras.  A total of 216 students- half of which were female- between the ages of 15 and 19 years old participated, with participation considered mandatory.  Students were told that in participating in the tasks at hand, they had a chance to earn money, and that one of the three sections would be randomly chosen for payment. Competitiveness, altruism, and risk taking were then measured for both genders.

The competition preference measurement consisted of three parts. In all three sections, participants were asked to engage in both a verbal and mathematical task. In the first section, the students were given two minutes to complete as many problems as possible. For each completed problem, each student was compensated with SEK 3 (about .50 US).  The second tier of the competitiveness experiment grouped the students into quartets. Students were told that if they completed as many as or more problems than their peers in each quartet, they would receive SEK 12 as compensation, but if they completed fewer, they would receive no money.  In the third part, the students had to solve problems, but this time they were given the choice to solve them independently or to compete against others.   After the competitive piece ended, students were asked to gauge their performance for both the independent tasks and the tasks in which they competed against their peers. This survey was used to define confidence levels of performance by students.

Altruism was measured by asking the students to participate in a dictator game, in which they had SEK 50 to distribute among themselves and a charity.  Finally, to measure risk preferences, students were given the choice between participating in a coin toss that would either give SEK 100 or no money at all, or taking part in a more conservative toss in which the award would increase incrementally in amounts between SEK 20 and 75. Each part of the experiment was followed by a survey.