American teenagers performed above the international average in a new test of collaborative problem-solving鈥攁nd 鈥渕uch better ... than would be expected based on their scores in science, reading, and mathematics.鈥
That was one finding of the Program for International Student Assessment鈥檚 first test of collaborative problem-solving, released Tuesday. When it came to applying their academic understanding to tasks that required complex teamwork, U.S. 15-year-olds outperformed those of some countries that had scored better than the United States in tests of traditional academic skills.
鈥淚t鈥檚 quite clear that individual skills do not guarantee that social collaborative skills will show up,鈥 said Andreas Schleicher, the director for education and skills for the Organization for Economic Cooperation and Development, which administers the PISA. 鈥淪ocial skills are not an automatic by-product of academic skills.鈥
Yet Singapore and Japan, two education systems that already outperform the United States in PISA鈥檚 math and science exams, also led the world in collaborative problem-solving. The United States ranked 8th of the participating systems.
Across 52 countries and economies that participated, only 8 percent of students scored at the highest level, showing the skill to not only identify paths and monitor progress towards solving a problem, but also staying aware of group dynamics, ensuring they and other team members fulfill their roles, and settle disagreements during the process.
Students worked through a series of tasks in multiple subject areas, both with real and virtual teammates, including problems that required teams to pool their information, negotiate, or arrive at a consensus. On average, U.S. teenagers performed at Level 2, meaning they could volunteer information and ask for clarification from team members, and sometimes suggest the next logical step to solve a problem, but they were less likely to be able to handle complex teamwork, group conflicts, or to evaluate the quality of information from different team members.
Interestingly, OECD also found some school practices led to less skill in collaboration. In science鈥攁rguably the subject in which students are most likely to have to work in teams鈥擮ECD found that students who spent time in the laboratory doing practical experiments or debating about investigations in most or all classes scored 31 points lower than students who rarely or never did so. Those differences held even after taking into account students鈥 academic performance, gender, and socioeconomic status.
That may back up prior research which has shown students are rarely taught explicit collaboration strategies when they are paired up in school, and that without such instruction, they often don鈥檛 really collaborate.
Different Gender Gaps for Individual and Group Work
Across all countries, girls significantly outperformed boys on collaborative problem-solving: 29 points higher on average worldwide, and 26 points higher in the United States. The gender gap remained at 25 points even after accounting for differences in boys鈥 and girls鈥 scores in reading, math, and science.
That鈥檚 a reversal from the OECD鈥檚 individual problem-solving assessment, back in 2012, in which boys outperformed girls. That test required students to identify information to solve a problem on their own, but did not require them to work with others.
So what鈥檚 going on? Schleicher suggested the gender differences may have something to do with differences in how boys and girls tended to approach collaboration in the first place. In a series of background questions administered with the test, the OECD found girls were more likely to value relationships as the core of collaboration, reporting, for example, that they 鈥渆njoy seeing my classmates be successful,鈥 and 鈥渆njoy considering different perspectives.鈥 Boys, by contrast, were more likely to agree with statements that focused on teamwork as a means to an end; e.g., 鈥淚 find that teamwork raises my own efficiency,鈥 or 鈥淚 find that teams make better decisions than individuals.鈥
鈥淭he first really value the nature of collaboration鈥攚orking with others, being a part, the empathic part of collaboration鈥攖he second basically just see collaboration as a means to an end,鈥 Schleicher said. 鈥淎nd that makes a difference.鈥
鈥淲hat鈥檚 interesting is this is not simply a reflection of social background. Even after accounting for gender and schools鈥 socioeconomic profile, you could see this pattern prevails. Attitudes and performance go together,鈥 he said.
Across boys and girls, students who valued relationships scored significantly higher than those who approached collaboration as transactional. And the highest-performing countries, Singapore and Japan, also had high percentages of both boys and girls who took a relationship approach to collaborative problem-solving.
Students鈥 broader social experience at school mattered a lot, too. Students who reported that they were generally well-liked by their peers scored 9 points higher on average in collaboration than students who felt unliked. And students who reported that their teachers had insulted them in front of others at least a few times in the last year scored 23 points lower on the test than students who said they had never or rarely been publicly insulted by a teacher.
鈥淭his suggests that the quality of student-teacher relationships is as important for learning how to solve problems collaboratively, as for acquiring knowledge and skills in science, reading, and mathematics,鈥 the study concluded.