Jaime Escalante
http://en.m.wikipedia.org/wiki/Jaime_Escalante
"In 1987, 27 percent of all Mexican Americans who scored 3 or higher on the calculus AP exam were students at Garfield High"
http://en.m.wikipedia.org/wiki/Jaime_Escalante
"In 1987, 27 percent of all Mexican Americans who scored 3 or higher on the calculus AP exam were students at Garfield High"
"Over the next few years Escalante's calculus program continued to grow but not without its own price. Tensions that surfaced when his career began at Garfield escalated. In his final years at Garfield, Escalante received threats and hate mail from various individuals.[8] By 1990, he had lost the math department chairmanship. At this point Escalante's math enrichment program had grown to 400+ students. His class sizes had increased to over 50 students in some cases. This was far beyond the 35 student limit set by the teachers' union, which in turn increased criticism of Escalante's work. In 1991, the number of Garfield students taking advanced placement examinations in math and other subjects jumped to 570. That same year, citing faculty politics and petty jealousies,[citation needed] Escalante and Jiménez left Garfield. Escalante found new employment at Hiram W. Johnson High School in Sacramento, California. At the height of Escalante's influence, Garfield graduates were entering the University of Southern California in such great numbers that they outnumbered all the other high schools in the working-class East Los Angeles region combined.[9] Even students who failed the AP went on to become star students at California State-Los Angeles in large numbers.[8]
Angelo Villavicencio took the reins of the program after their departure and taught the remaining 107 A.P. students in two classes for the next year. 67 of Villavicencio's students went on to take the A.P. exam and 47 passed. Villavicencio's request for a third class due to class size was denied and the following spring he followed Escalante and quit Garfield. The math program's decline at Garfield became apparent following the departure of Escalante and other teachers associated with its inception and development. In just a few years, the number of A.P. calculus students at Garfield who passed their exams dropped by more than 80 percent. In 1996, Villavicencio contacted Garfield's new principal, Tony Garcia, and offered to come back to help revive the dying calculus program. His offer was rejected.[8]"
While he was rather strict/demanding, too, the article also hints to one of the more important causes of his success: "He rejected the common practice of ranking students from first to last and instead frequently told his students to press themselves as hard as possible in their assignments." In educational science the former is called social reference standard and the latter individual reference standard. While sadly the former is still in use more often, longitudinal studies have shown a strong positive effect on achievement for individual reference standards (I only have a German source right now: Rheinberg & Krug, 1999). The reason seems to be that students seem to attribute ratings with social reference to stable attributes like talent ("I'm just not good at math"), whereas individual benchmarks are more often interepreted as a result of (lack of) effort. Interestingly the socially disadvantaged have shown the strongest results with individual reference standards, just as in the case of Jaime Escalante.