Correction

The paragraph discussing the NSS should have said that teaching and feedback are derived from undergraduate students during their final year at university, not after they have graduated. This was corrected on May 1.

Today's Guardian university tables are not the last word in judging the quality of UK universities; they are a starting point for students looking for a university course.

We know they attract an enormous amount of interest in Britain and around the world. They are also unpopular in some quarters, of course - but we'll come to our academic review group later.

This year, for the first time, we have used data from the National Student Survey (NSS) to get a sense of what students think of their experience at university. That has produced a lot of changes since last year, and means the tables are not strictly comparable with previous years.

Our aim is to provide a guide for first-time students, not to assess the general standing of universities. That means we concentrate on teaching and do not include research ratings, which count heavily in other league tables. A postgraduate guide might look very different.

It also means that we delve deeper into data for particular subjects. We believe that this is of more use to students preparing to apply; no matter how good a university's reputation in general, it's not much use applying if the subject you want to study isn't taught well there.

We assume young - and not so young - people who are capable of doing a degree are capable of using our tables as a starting point before investigating the style and content of particular courses, the more general attractions of a university, campus versus city and so on.

We chose the title the Guardian University Guide 2008 because most of our readers will be looking to start courses in September next year, although we hope the tables will also prove useful in the stressful August days of clearing when students need clear, comparative information in a hurry.

The tables are based on data for full-time undergraduates; we haven't yet cracked the more complex task of measuring part-time students, so distinguished institutions such as the Open University and Birkbeck College do not appear. One day, perhaps.

We have rated courses against the following criteria:

· Teaching quality, as rated by graduates of the course

· Feedback, as rated by graduates of the course

· Spending per student

· Staff/student ratio

· Job prospects

· Value added - comparing students' degree results with their entry qualifications

· Entry score

From these we have compiled an average for each institution, weighted by numbers of students and mix of subjects.

If you don't like our selection you can go to EducationGuardian.co.uk/universityguide2008 and create your own tables by giving more weight to, say, job prospects or spending per head. Students are not all looking for the same things, after all.

The tables have been compiled for Education Guardian by Campus Pi, an applied research department at Brunel University. The rankings are based on official information published on universities and HE colleges collected by the Higher Education Statistics Agency (Hesa) and the NSS published by the Higher Education Funding Council for England (Hefce).

Teaching and feedback are derived from the NSS, now in its third year and generating a high rate of feedback from undergraduate students who are questioned during their final year at university.

Scottish universities are coming on board, though only three made returns for last year's survey, on which our tables are based. In England, Warwick, Oxford and Cambridge are also expected to follow suit. In the meantime, universities that do not complete the NSS, or departments where the returns were not high enough to be reliable, have been given a score in line with their other criteria (spending, entry score and so on).

In judging teaching, students are rating their experience against expectations, but universities that generate high expectations do not seem to come off worse.

On the other hand, there are differences between subjects when it comes to student satisfaction. Art students, for instance, seem to be grumpier than historians or biologists.

Spending and staff/student ratios are compared only between departments teaching the same subjects - medical students obviously attract a lot more money than those in business studies.

The scores in the tables are expressed in bands - one to 10 - and the "bandwidths" will be published on the website for those administrators and academics who want to check the exact figures. The banding calculations have, we believe, been improved; a more detailed methodology is published on our website.

Feedback is the other NSS-derived score; we know this is a big issue for a lot of students and the aspect of their courses that they rate the lowest. We don't accept the jibe from some academic staff that students complain about feedback when they get bad marks. It's an area in which universities are conscious they need to improve. When a prestigious university department is getting a high score for teaching and commands high entry scores but gets one out of 10 for feedback, then alarm bells should be ringing.

League tables have been blamed for degrading university standards in the pursuit of ratings; this is to be investigated by Hefce and we wait to be summoned to justify our efforts. But if tables induce universities to pay more attention to teaching and feedback to students, they may do some good.

Inevitably, the main league table gets a lot of attention, but the striking thing about it is how little separates so many institutions. There is not a lot between Oxford and Cambridge at the top and even less between St Andrews and University College London, almost tied for fourth place behind Imperial. Edinburgh and Warwick are neck and neck, as are Bath and Loughborough. (St Andrews has been propelled up the table by very good student satisfaction scores, as have some of the other risers.)

The institution table does not include specialist colleges offering only one or two subjects, but in their fields they can shine. In music, for instance, the top four spots go to Trinity Laban, the Royal Academy, the Royal College and the Royal Scottish Academy.

Conscious that students (and their parents) are concerned that their investment of time and £3,000 a year will make them employable at the end of their course, we have included an online guide to the skills taught in various subjects, compiled by universities and employers.

Psychology graduates, for instance, should develop planning and financial awareness as well as analysing their complexes; materials engineering teaches teamwork and communication as well as how to manufacture stuff.

Two universities - London Metropolitan and Keele - declined to let their data be used in our table. We are grateful for the cooperation of those that have and thank them for their patience, though we do try to minimise the hard work by using data they have submitted to Hesa anyway.

Discussion has at times been robust, but always constructive, and to see if we can improve the accuracy of the Guardian tables and smooth the process of compiling them, we are planning to convene a review group drawn from the universities to help us.

· If you have any comments or points for discussion, email them to education.editor@guardianunlimited.co.uk. To create your own tailor-made table go to EducationGuardian.co.uk/universityguide2008