"When you see somebody with it, you know they have it," says American television star and opinion-shaper Oprah Winfrey. The ‘it’ factor she says, is "an undefinable quality," a light that "cannot be dimmed, no matter how much hardship or poverty you have known." Yet that intangible quality is exactly what Oprah believes she was able to identify among the successful applicants she interviewed for a place at a $40 million girls boarding school she has funded and is championing in South Africa.

Located on 52 acres near Pretoria about 30 miles south of Johannesburg, the Oprah Winfrey Leadership Academy is described as 'state of the art'. Out of 3,500 applicants, her school has admitted just 152 11- and 12-year-old girls, all of them poor and black, and Oprah has built herself home on the campus.

Oprah has been criticised in the US for investing in South Africa when so much unmet need exists in her home country. The talk show host was also reported as having unfavorably compared the aspirations of American poor families with those of their African counterparts. Merged into the coverage of the hoo-hah surrounding the school's star-studded opening, these stories mask two more important issues for prevention science.

In all countries of the world, living away from home is a rare experience and going to a boarding school is rarer still. There are no reliable data, but probably no more than 200,000 of the US child population of 90 million are in boarding schools. Also to be reckoned with is the damning evidence about the value of some of them: the residential schools which many indigenous children in Canada, Australia and the US were once forced to attend have left a bitter legacy. Nevertheless, the promise such places seem to hold for solving entrenched disadvantage has captured the public and philanthropic imagination for generations — and does so still.

Oprah is not the first American mogul to invest in residential education. In 1907 Milton Hershey donated his wealth to build a residential school for poor white children near to his chocolate factory in Pennsylvania. More recently, relying on public and private fund raising, Eric Adler and Rajiv Vinnakota established the Seed Schools in Washington D.C. Another initiative is testing a boarding school programme as a long-term placement option for foster youth with little prospect of family reunification.*

The roots of such enterprises might be traced back to the building and aristocratic endowment of boarding schools in 16th century England. Some of their modern counterparts such as Hershey and the Kamehameha Schools in Hawaii, now struggle to use their huge endowments effectively. But for all the long history of residential education and the enduring interest in it, no rigorous evaluation has attempted to assess its impact on child outcomes or to compare it with other forms of provision.

Oprah Winfrey’s talent and determination are undeniable. She survived an impoverished childhood after being thrown out by her mother and being raised by a grandparent. Her commitment to the Leadership Academy goes far beyond donating money. In addition to funding and personally meeting and interviewing each of the shortlisted group of students, she visited some of the candidates in their own homes. The selection criteria included academic ability, demonstrating leadership potential and living in a family whose income is less than $750 per month. But the final criterion was that ‘it’ factor which only she could discern.

This striving to identify students who have most to gain from residential education is common among legacy schools, and, for all her gifts and good intuition, Oprah is unlikely to be any better at it than anyone else. Attrition rates from most boarding schools, for poor or rich, run at about 10 per cent per year, meaning that for every 100 students enroling only about 50 will graduate.

Reporting of the creation of the Leadership Academy also overshadowed the creditable performance of the South African education system as a whole. The United Kingdom has a GDP per capita roughly two and a half times greater than South Africa, but high school completion rates in the two countries at about two-thirds are roughly equivalent. Even if Oprah's school is a success, it seems unlikely that a similar investment could be made for all of South Africa's children.

Like most legacy schools of its kind, Oprah's provides the perfect opportunity to discover the impact of residential education on children's health, education and development. ‘It’ factor apart, selection criteria are limited and easily measurable; potential beneficiaries far outnumber spaces available. By randomly selecting students from the pool of eligible applicants, and comparing the progress of those admitted with those turned away, not only could Oprah serve poor children, she could also help the world understand who benefits from this kind of intervention, when and why. By personally selecting the students, Oprah has bypassed an opportunity to learn, at least for now.

* See Jones, Loring and John Landsverk, "Residential Education: Examining a new approach for improving outcomes for foster youth," Children and Youth Services Review, 28(2006), 1152-1168