I once worked with a great head of school at a very good school. In fact, it was the school at which I cut my teeth and developed some of my foundational educational philosophies. I asked the head why, in spite of high student demand, we were not more selective when admitting students. After all, we could afford to accept only the top academic performers.
His answer stuck with me: “Keith, any school can take in cream and put out cream. I am interested in taking in milk and putting out cream.” Excusing the dairy reference, his motives were clear. Really good schools took students at varying levels and brought them all to a much higher level by graduation.
People often focus on standardized test scores such as IB, AP, SAT, and other single measures of performance to demonstrate success. Research confirms that top-testing students generally continue to perform well in college. If you want to fill your class with students most likely to do well in college, you'll pick those with the highest high school grades, and mostly females. These statistics do not necessary predict how graduates will perform later in life, but they are statistics all the same. However, as mentioned, these are single measures of student achievement and do not reflect motivation, effort, resilience, leadership, or moral goodness. Many educators would argue they do not even truly represent academic achievement.
"Setting", "streaming", or other nomenclature designating a process by which schools separate students by ability/talent/achievement, is hotly debated. If you think about it a bit, it does make some sense: take the top students in math and have them do math with other top students. They should achieve at a higher rate than the norm. And of course, that also means separating out the students not as strong in math and expecting them to achieve at a lower level.
Does it actually work that way? Large-scale research demonstrates that is not the case. Setting or streaming actually has a minimal effect on achievement, including for top students. However, it does have deleterious effects on the other students. (J. Hattie, 2009)
Here's a short story, and then I'll get back to the research.
My best friend from school was a streamed student. He was put into lower level math classes, while I was in the highest level class. In fact, in his junior year of high school, his class was often called (even by the teacher!) the “dummy group.” Not surprisingly, he learned to hate math and believed he was not good at it.
Fast-forward 20 years: After studying animation in college, he went on to start a small animation company, while I went on to study math and engineering. I became a math teacher, and eventually a headmaster. He discovered that his poor ability in math was not such a problem after all -- in fact, he managed tight budgets, convinced investors to lend him money, and negotiated outstanding contract terms. He managed a payroll that employed 400 people at its peak. He eventually sold his company after accurately evaluating its net worth, and retired at 48. He loved his work and realized late in life that his math abilities were not as poor as he was told. On the contrary, he was a brilliant businessman.
My point is this: We live in a world in which we must work with people of varying abilities and aptitudes, who come from different backgrounds and cultures. My friend had the ability to work with hundreds of artists (no simple feat) and create a great company through his fiscal intelligence and interpersonal skills. Let us not decide on someone's aptitudes too early.
Our job as educators is to expose young people to as much diversity as possible; help them learn to work with all types of people in a variety of situations; and encourage them to be open-minded, take risks, and develop confidence. This is perhaps more important now than it has ever been.
Separating young students by “perceived ability” has many pitfalls and does not often nurture the type of people who will make the most difference in the world; I think Bill Gates and Warren Buffet would agree. Remember that Albert Einstein made some of his greatest discoveries while employed as a patent clerk, before he became a faculty member at Princeton.
Back to the research! It's been shown that resisting the labeling of students has a significant positive effect on student performance, while setting or streaming provides very small positive effects. In fact, other strategies with similarly minimal positive effects include small class size, homework for elementary students, single-sex schooling, and holding students back (retention). Large positive effects result from things like teacher quality, phonics, small group and collaborative work, and direct instruction. It is not a coincidence that TASIS employs these positive strategies while avoiding those not supported by research.
In short, we believe in quality teachers, trained to differentiate instruction, who provide challenge to students who need more and support to those who are struggling. That's the best learning environment to produce the best human beings.