Questions of context: low marks for crude admissions policies

As a subscriber to Times Higher Education, I've been surprised by how few of its articles have been written by representatives of the independent schools sector. After all, we have much in common with higher education: we both value scholarship; we both charge fees; we both select students; and we are both interested in identifying talent, wherever it is to be found.

Moreover, contrary to what many may believe from newspaper headlines, most independent school heads are not opposed to the use of contextual information in deciding who attends university. In fact, they use it as part of their own admissions processes. Nor are they against social mobility - quite the opposite, as proven by their growing use of means-tested bursaries. Our sector supported the 2004 Schwartz report's main proposals on fair admission to higher education.

So why are we concerned about university admissions? There are three main reasons. First, the growing politicisation of this issue. In spite of repeated official statements that the Government respects university autonomy and does not support any move towards quotas for under-represented groups, various ministerial speeches (and comments by Gordon Brown on David Cameron's schooling) have stirred up memories of the Laura Spence affair and led to headlines about "class war". It is, of course, perfectly appropriate for independent think-tanks to suggest that if universities want to achieve a better social balance they should engineer such an outcome. But if the Government wants to engage in social engineering of this sort (and some would say this is a legitimate aim), it should at least be honest and say so. The first Schwartz principle, transparency, cuts both ways.

The second issue is a lack of linguistic clarity. Although the Higher Ambitions framework document marks a welcome break with the past in only occasionally using school type (maintained versus independent) as a proxy for disadvantage, most official documents seem incapable of deciding what they wish to combat or compensate for. Students are referred to as "non-traditional", "disadvantaged and gifted", "from the poorest homes", from "lower socio-economic groups" or "less privileged backgrounds". How are universities supposed to know which of the many indicators of disadvantage they should use?

Schwartz distinguishes between "soft data" (socio-economic indicators such as free school meals, educational maintenance allowances, postcodes, parental background and so on) and "hard data" (in particular, state data relating to school and college performance).

The latest manifestation of the latter is the so-called "modifier" in which school or college type (this time "high performing" versus "low performing", as measured by government performance tables) is used to inform judgments about applicants' educational contexts. But as Juliet Chester and Bahram Bekhradnia point out in the Higher Education Policy Institute report, Oxford and Cambridge: How Different Are They?, it is hard to disentangle this from other factors that affect educational achievement, such as parental background.

Modifiers, such as those used by Durham University, are blind to the fact that a student on a bursary in an independent school may come from a low-income family. Equally, a student from a poorly performing comprehensive may benefit from private tuition at home. At present, modifiers are very crude tools, unfit for purpose.

The final concern relates directly to this. Without proper research, including pre-testing, piloting and evaluation, such "indicators" are highly suspect. Even relatively tried-and-tested devices such as postcode data or free school meals are known to have limitations. Doubts about modifiers relate both to the reliability of the data on which they are based and the validity of using "school context" as a proxy for "educational context".

What then, in the short term, are admissions tutors to do in trying to put students in context? The simplest answer is to use a "basket of indicators" that are qualitative, not just quantitative - what Supporting Professionalism in Admissions calls "holistic assessment". Within this, the "school reference" must be a key source of validated contextual data. Heads in the maintained sector see the new School Report Card as potentially giving a much more valid picture of the overall context and achievements of a school than either the raw or value-added data emanating from the Government.

If all else fails, of course, there's always the admissions lottery. As one vice-chancellor argued persuasively, it's one of the few tools available to universities that can't be challenged in court. Asked to choose between a die and a modifier, some applicants, I suspect, would choose the former. Most want neither: they just want a fair system of admissions that treats them as individuals, not arbitrary numbers.