Why participation isn’t widening

Frustrated at my ongoing indisposition – I had to miss today’s Admissions Day at Sussex University, which has put me in a very bad mood – I’ve decided to deliver a short rant about widening participation. WP is the name given to schemes to open up access to higher education to students from less advantaged backgrounds. An excellent idea, of course.

The problem is that, despite pressure from the relevant quango (OFFA) most self-styled leading universities, especially those in the Russell Group, have consistently failed to widen participation to any significant extent. Why is this?

The easy answer is that universities have to take students who are adequately prepared for undergraduate studies, which means selecting on the basis of A-level grades, which means students from private schools have an advantage.

The problem with this argument is that, at least in Physics and Mathematics, I don’t think A-levels are a reliable indicator of aptitude for undergraduate study at all. If I had my way we wouldn’t use A-levels at all.

Unfortunately we’re stuck with the current, unfair, system because any “leading” university that takes a large number of students with weak A-levels (possibly through a Foundation Programme) will be penalised in the league tables for not being selective enough. Moreover, the Government’s decision to lift the cap on places for students with AAB or better, means that recruiting students with top A-level grades is potentially the most lucrative strategy.

That the system doesn’t work the way it is supposed to is obvious. If you don’t agree, then ask yourself why it’s not the case that virtually all Oxbridge students get first class degrees, when they admit only A*/A students?

Things won’t improve until we abandon the obsession with A-level tariff points and find a way of assessing intrinsic ability.


30 Responses to “Why participation isn’t widening”

  1. David Howey (twitter @davidhowey) Says:

    So…… How to assess intrinsic ability then? we have a physics aptitude test in addition to a levels for example

    • telescoper Says:

      Use the aptitude test but not the A-levels!

      • From my experience in Oxford you may also need to include information on the type of school. In many subjects (especially the physical sciences) state schooled students getting into to Oxford achieve a significantly higher proportion of firsts (similar 2:1 proportion and lower 2:2 3rd proportion) than their independently schooled peers.

    • Bryn Jones Says:

      Hold on.

      Students need several qualities to cope successfully with a university course. These include: (1) an aptitude for study in general and and aptitude for a subject in particular; (2) an understanding of the basics of a subject including of necessary techniques; and (3) a willingness to work. A-level examinations are supposed to test all these. It’s not clear to me that aptitude tests could do so.

      • Of your list, I think A-levels can judge 2 quite well, and 3 and 1 less well, in order of decreasing effectiveness. However, for most science/maths degrees you need the basics before you can move on to the real meat of the subject. It’s bad enough tat we have to spend much of the 1st term of a physics degree in filling in the gaps between different syllabuses. If you had to start from scratch, it would be difficult and might add another year to the degree.

        Furthermore, any test can be taught, so those taking an aptitude test with extra (payed for, whether through tutorials at home or at a fee paying school) tuition will appear to be better candidates.

  2. Garret Cotter Says:

    Steve, so you have a reference for that? According to Ogg et al., 2009, BERJ, 35,5 “school effects are only statistically robust for arts students”.

    • Garret: I’ve emailed you the document I was sent. As you can see this is far from a finished study and could be deely flawed or biased, it would be great to do our own analysis.

      However I did make a mistake, the number of firsts is comparable however the ratio of 2:1s and 2:2s apears quite different.

  3. Anton Garrett Says:

    I don’t believe that Russell Group Universities are deliberately taking rich thickos just *because* they are rich. So it is totally unfair to penalise them. They simply differ, in good faith, with others about how best to measure aptitude. Some universities used to have their own entrance exams, but those got abolished mainly because they were consuming of academics’ time (setting and marking) at a time when government was increasing the amount of admin falling on researchers.

    Optimal solution is of course to improve State school standards so that there is less need for universities to have to estimate the future of their applicants.

    • The day that state schools would catch up with private schools would be the day that private schools would die off, as their reason for existence would be largely undermined. After all, they don’t really sell education, but privilege, and the danger with universities is that they also end up selling privilege instead of fulfilling the needs of the society through educating people.

      Entrance exams are the fairest way of testing applicants, if people need to be tested. Beyond that, there’s a very easy way of balancing student intake: Restrict the number of students from private schools to their percentage of their age group in any university that takes any money at all from the state. Most universities would become far more representative and parents who could pay for private schools could continue to pay their children’s education through increasing number of new private universities.

      • Anton Garrett Says:

        If you have entrance exams then a student applying to three universities must take three entrance exams, unless the universities get together and institute a common entrance exam. In which case why have A Levels?

        State schools could be made a LOT better if basic English were taught via phonics, mathematics taught far more by rote in its early stages and without too many examples (eg 12/3 not 12 pence divided among three persons), meaningful discipline could be exerted on unruly pupils, and it was easier to sack poor teachers. (Like the teaching unions I do not approve of performance-related pay; just make sure that the teachers you have are good.)

      • I certainly agree that improving the state sector to match the achievement of independent schools would greatly reduce the influence of the latter. There are places where state schools are uniformly poor, and so aspirant parents with the resources will send their children to independent schools. I would hesitate, though, to force the formation of private universities since that will just push the problem up in age. The private (and, if the US is anything to judge by, well resourced) universities would soon become the reservoirs of privilege and maybe better university education as well, especially if they start giving scholarships to the better state school children (again, like the US).

        It can be argued tat the only reason certain sections of the UK support state funded UK universities is because there is no non-state funded alternative.

      • Anton: The ‘Cambridge PreUs’ are already accepted as an alternative to A-levels by other universities.

        [It used to be the case that you would have to take A-levels and both Oxford and Cambridge exams when they existed (though I don’t think many applied to both as the lower rated Oxbridge university would almost certainly reject the candidate) but that is many years ago.]

        My own solution is to make A levels more stringent so that it is easier to spot the top 5% of candidates who are the ones the top universities are after. It used to be this way when A grades were based on relative not absolute performance. You could add a percentile to grades today to achieve both, but A level marks, at least in maths & physics, too often saturate with a lot of the candidates I see frequently getting marks above 95%.

        At the moment we have to interview everyone who looks good, using a huge amount of staff time. And this is necessary since I’ve found some truly bad candidates in spite of their high marks in AS and projected in A2s (and from both state and famous-name independent schools).

      • Anton Garrett Says:

        The better State schools USED to give an education entirely comparable with private schools, but something went wrong in the last 50 years.

      • Anton: From what I hear from friends with children, it is still the case that the (very) best state schools are comparable to independent schools for quality of education. My experience on admissions would tend to agree with this.

        The problem is the increasing hinterland of state schools that are not that good, and the other factors behind the establishment’s abandonment of state schools. At least we don’t have a private university system in the UK that would allow for entirely separate development.

    • Fundamentally I don’t think there is any easy or accurate way of determining which students are going to thrive at university.

      My argument here is that the learning environment in schools is very different from universities (this difference is perhaps wider for the independent sector due to the greater access to small group tuition).


  4. Andrew Liddle Says:

    The challenge here is distinguishing aptitude from preparedness. Private schools seek to to deliver the latter and I suppose do so successfully, while the A-levels assess some combination of the two (as would any entrance exam). I suppose a cleverly constructed entrance exam *in combination with A-levels* might allow some sort of decomposition of a student’s ability into aptitude and preparedness. I don’t think it can be right to ignore the A-levels completely as they are clearly conveying some sort of information.

    WP is supposed to target students who are high on aptitude but have been significantly disadvantaged on preparedness. Unfortunately, as Peter says, universities tend to be penalized indirectly for taking on such students, through lower league table positions which ultimately affects both income and the perceived status of the university. In some league tables this is supposed to be accounted for through an `added value’ driver, which presumably compares final degree class to entry grades, but its workings are unclear.

    Universities are also reluctant to take on students with a wide range of preparedness because it implies running different levels of courses, such as the Foundation Year model (which has been a great success for Sussex over the years with many students eventually achieving first class degrees). But the WP funds made available to deliver such programmes have never been adequate, hence the `elite’ universities that don’t need them for financial reasons tend not to have them.

    And of course the current government’s HE policies are a massive step backwards against the modest progress that was being made, in particular putting further massive financial pressure on students seeking to following the Foundation Year routes.


  5. Will Dennis Says:

    “That the system doesn’t work the way it is supposed to is obvious. If you don’t agree, then ask yourself why it’s not the case that virtually all Oxbridge students get first class degrees, when they admit only A*/A students?”

    I’m sorry, but you can’t cite this as some sort of proof that the A level system is somehow completely broken and should not be a consideration in university admissions. For one thing, as you know universities set and mark their own exams internally, which leads to wild variability in the amount of content, difficulty of exams and general standard/threshold in which to get a required grade. Let’s take Maths as an example, as the primary entry criteria at Cambridge are the very challenging STEP papers (aimed at the top 2% of maths students) rather than A levels. I think it’s fair to say that in general, the Cambridge student cohort for maths is going to be a lot, lot stronger than a mid-ranking university like, say, Aberystwyth. But it seems, they award the SAME proportion of first class degrees (30%). So either the staff at Cambridge are so awful that despite having a stronger set of students they end up the same standard after three years as a set that were much weaker, or the initial standard of each cohort was the same, or in fact the courses were of very different difficulty and tailored to be equally challenging to their respective cohort.

    A quick perusal of the syllabus and past papers for maths from both Cambridge and Aberystwyth strongly suggest the latter to be the case: a typical Aber first year paper (http://www.aber.ac.uk/en/media/MA1002c-11.pdf) and a typical Cam first year paper (http://www.maths.cam.ac.uk/undergrad/pastpapers/2011/ia/PaperIA_1.pdf). You can find the full set of either fairly easily. I don’t really think it’s fair to say that a student who achieves a first class mark in either of these two courses are absolutely equal, in fact it’s somewhat insulting and degrading to suggest this. If anything, it’s the broken system of external examining that needs fixing. I would absolutely guarantee that if all university exams were centralised and all students made to sit the same papers, then you probably WOULD see Oxbridge students getting virtually all first class degrees. I myself struggled greatly at Cambridge studying physics, often finding myself around the 2:2/2:1 boarder, until transferring to another institution where I found attaining a high first class almost farcically easy.

    If A levels really were such an awful indicator and complete waste of time then why ARE universities using them to select students? Well, Cambridge have actually done some research into this and it actually appears that there IS a link between A level performance and how well you do on your degree. (http://www.admin.cam.ac.uk/offices/admissions/research/a_levels.html) Reading the study, it seems that they are the strongest indicators for *gasp* NatSci, which includes Physics!

    Now, obviously A levels aren’t perfect, and yes, some excellent students don’t do well in them for whatever reason and are in fact very well suited to degree level study, but they do at least suggest that on the whole, the better you do in your A levels, the more likely it is you’ll do better in your degree.

    • telescoper Says:

      I’m afraid your comparison between first year papers is irrelevant because these don’t count towards the degree classification. Second and third year papers are moderated by external examiners to ensure consistency. The final-year physics papers from Cambridge I have seen are no harder than those from other institutions, although they do have peculiarities in the way they are marked.

      • J.C. Whitehead Says:

        I’m rather surprised by this answer. Surely even if the questions are comparable, the various grade boundaries could vary from institution to institution?

        As far as maths goes, the explanation of Will Dennis’ observations cannot simply be that first-year is a special case. He might just as easily have pointed to a pretty astonishing second-year numerical analysis paper, http://www.aber.ac.uk/en/media/MA25110-11.pdf ; or a second-year linear algebra paper, http://www.aber.ac.uk/en/media/MX31410-11.pdf : both compare unfavourably with the questions asked of second years at Cambridge in the same year, available at http://www.maths.cam.ac.uk/undergrad/pastpapers/2011/ib/List_IB.pdf .

        Surely this – frankly astounding – disparity suggests whatever system of moderation is in place is failing?

      • telescoper Says:

        Perhaps so, at Aberystwyth..

        …although it’s hard to say without seeing the mark scheme for the Cambridge papers.

      • J.C. Whitehead Says:

        There’s really nothing special about Aberystwyth; dozens of other universities have the same rather pedestrian examinations. If it’s failing there, then it’s also failing at each of the others – and hence failing in general. Sussex has its papers hidden behind an authentication barrier, but I imagine they’d be similar to, if perhaps marginally tougher, than Aberystwyth’s.

        The reality is that maths students could score 100% on the Aberystwyth paper – or those at other similar institutions – without having the first idea of how to tackle the challenging questions on the Cambridge paper, which require a measure of independent and original thought, and knowledge of a far wider syllabus (both the sheer size of individual courses bearing the same name, and the number of courses taught is greater at Cambridge).

        Full disclosure: I’m a second year maths undergraduate at Cambridge, and I am absolutely convinced that the system of external examiners is failing. There are students here who will come out of their degrees with 2:2s, unable to get onto graduate programmes at universities and companies, who will have surpassed by some margin the (demonstrated) mathematical ability of first-class graduates of other universities (like Aberystwyth and others). Fortunately, with a good measure of hard work, I don’t think I’ll be in that position, but I have friends who might. This is abysmal; the system is failing, and needs reform. That is the conclusion I draw from the similar proportion of firsts awarded by different universities – not that the admissions selection system (STEP – a fantastic idea) is flawed.

        Incidentally, the Cambridge maths marking system is obscenely complicated and rather opaque – but in rough summary, to get a first you’d need to complete perhaps 15 of the long questions and 5 of the shorter questions to a high standard (ie essentially entirely correct) overall. You are penalised quite heavily for incomplete answers – the “bonus” of 30 points becomes a bonus of only 5. There’s lots of information about (just google “alpha beta cambridge maths examinations”).

      • Bryn Jones Says:

        Those examples do suggest that there are problems with standards with some mathematics papers at Aberystwyth. I do not know how that has come about. I am not familiar with standards in other subjects at Aberystwyth to comment on whether other subjects are affected, and certainly have no evidence that they are.

        Aberystwyth University has grown out of the old University College of Wales, which was part of the old University of Wales. Historically, from the late 19th century through much of the 20th century, the University of Wales attempted to match its examination standards with those at other British universities, and explicitly with standards with Oxford and Cambridge.

        I disagree very strongly that standards at Cambridge are superior to other universities. There are certainly fundamental differences in standards across the British university sector as a whole, but these are between the “top end” and the “bottom end” in general. I do not accept there are any substantial differences between, say, Russell Group institutions, and I did not see much difference during the teaching I did in a few of those. The mixing of external examiners between these institutions should maintain comparable standards in examinations, including of grade boundaries.

        I am very aware that Oxbridge undergraduates are told about the superiority of their institutions, particularly within colleges, but I am unconvinced there is evidence of this. Certainly, it is difficult to see differences in ability between Oxbridge graduates and graduates of other universities with similar degree classifications. Oxbridge graduates tend to be outwardly confident and talk very well, but there seems no noticeable difference in terms of the actual ability behind this.

      • J.C. Whitehead Says:

        I think it’s a little unreasonable to suggest that my conclusion is somehow born of a smug institutional arrogance – nobody at Cambridge has ever told me it is superior to other universities, or the course harder. I have come to that conclusion, for my subject only, after looking at other universities’ exam papers and seeing their grade statistics.

        I think it’s difficult to generalise to other subjects in part because no other subject has the admissions hurdle of STEP, which changes the field enormously, and makes Cambridge mathematics something of a special case – no doubt why Will Dennis used it as an example in his comment.

        I agree entirely that other Russell Group universities have courses much closer to mine in their degree of difficulty – but the external examiner system is failing if there is a marked difference in standards between any two universities, not just two in the same prestigious subset.

        But I’d like to emphasise that Aberystwyth is not a special case; this is not an isolated discrepancy. Sticking to second-year linear algebra, Leicester ( http://www.math.le.ac.uk/TEACHING/UG/PAST_EXAMS/MA2102_Linear_Algebra_2002-2003.pdf ) and Liverpool ( http://www.maths.liv.ac.uk/Past_Exams/PDF_FILES/MATH244-may07-exam.pdf ) both provide reasonable examples of papers harder than Aberystwyth’s but still far more pedestrian than Cambridge’s.

        I don’t personally derive any pleasure from this observation – rather, it suggests I could get a better degree for less effort were I studying elsewhere, and that, although (with less effort and easier exams) I would likely end up worse at mathematics, those who believe standards are consistent across the country would believe me better! Perhaps my friends at sixth form college who were put off applying to Cambridge because STEP seemed too challenging had the right idea after all.

      • Will Dennis Says:

        I have to agree with J.C. Whitehead. It is somewhat patronising and insulting to be told the only reason you believe that a Cambridge course might be harder is because your institution tells you so. It may well be that once upon a time Aberystwyth set difficult maths exams comparable to Cambridge’s, but in these days where ~40% of a given year are going to university and league tables are constantly published with ‘% good honours’ as a key criteria this has inevitably lowered standards. It makes no real sense to give Aberystwyth students papers as hard as Cambridge papers – it’s likely that most would not do very well and the exam would not be very discriminatory to that cohorts’ ability, all the marks would be clustered at the low end.
        Similarly to J.C.s observation, I infact took it upon myself to leave the Cambridge NatSci course for another (Russell Group) university, and my grades went from high 2:2s to very high firsts (with far less effort I might add). On the face of it I’m now one of the best Physics students in the UK despite being very average indeed within the Cambridge cohort.
        I think the past paper comparison can be extended to more subjects than just maths, the reason I picked that subject was that it admitted students based on more than just A level grades. Looking at the difference between Sheffield’s fourth year physics papers and Cambridge’s, there is a clear difference in the style and difficulty of the papers, not to mention that the material examined (at least as far as Condensed Matter and Semiconductor Physics is concerned) is much more advanced; I could not even find a Sheffield paper that examined much of the Cambridge material, and this is considering Cambridge students study other subjects in depth in their first and second years! (http://physics-database.group.shef.ac.uk/exampapers/2011-12/PHY410-11-12.pdf) (http://www-teach.phy.cam.ac.uk/dms/dms_getFile.php?node=7273)
        Thankfully not all academics seem so quick to dismiss this, and an interesting document produced by Bill Lionheart at Manchester (http://www.maths.manchester.ac.uk/~bl/wheretostudymathsuk.pdf) actually states ‘There are likely to be people who would have gone on to do research in mathematics had they gone to an institution requiring lower grades’ and ‘Overall the best strategy is likely
        to be a compromise between optimising your expected degree classification and the reputation
        (deserved or not!) of course and university where you study.’

        It seems crazy now that the best advice for very good sixth form students is now ‘go to where the course is easiest’ rather than ‘go where you will maximise your ability at your subject’ due to this ridiculous failed system of degree classification.

      • Bryn Jones Says:

        There perhaps is some irony that I’m the person accused of being patronising when I do not talk down the standards of teaching in the very best universities in Britain, and instead defend the work of academics teaching in them.

        The central problem regarding standards is not claims that Cambridge has higher standards than Imperial College, or Oxford than UCL. It is that there are differences in the standard of the assessment between those several institutions regarded as the best in Britain, and the institutions commonly found 150 places down university league tables, whatever in reality those league tables measure.

        To answer Philip’s point, I do not accept that there no differences in standards across the British university system. I believe there are differences, big differences, but that the external examiner system maintains standards across numbers of institutions of similar standing. The institutions that Philip knows for their research in astrophysics will mostly be found close to the top end in terms of quality, and the external examiner system will ensure standards in those instutitions will not differ enormously.

        The scandal is that students who pay £9000 fees per annum to attend Chigwell Metropolitan University or Scunthorpe City University (to give two hypothetical examples) receive degrees that potential employers regard as wholly inferior to those from the institutions regarded as of leading quality. Many candidates apply to university without understanding this. And the people who suffer worst are those who come from families with little university participation, who do not get the advice about which institutions to apply to.

        There needs to be independent external assessment of degree standards across all British universities. That would help to standardise the system. And I am confident that little adjustment would be needed between institutions at the top of league tables – whatever those tables measure. There would need to be significant adjustments to make the lower end comparable to the top end. How such a system could be constructed in practice is anybody’s guess.

    • J.C. Whitehead Says:

      I appreciate that this discussion has lain dormant for quite some time, but I recently came across some interesting relevant material – and, the rather hostile reaction my observations generated having rankled with me since, thought I would come back and share it.

      The Maths Faculty at Cambridge has a Teaching Committee whose responsibilities include the analysis of each year’s examinations (at all levels of the undergrad Tripos). The reports of the last ten or so years contain quotations from the external examiners – many of which suggest not only that Cambridge’s examinations are more exacting than those elsewhere, but also that the external examiners at Cambridge see no problem with this disparity.

      Rather than repeat the points I made above, I will let you draw your own conclusions. Obviously I do not expect these to extend immediately to other subjects.

      The reports are available here ( http://www.maths.cam.ac.uk/facultyboard/teachingcommittee/ ), and I’ll give a few especially pertinent examples:

      “The breadth of topics covered in the exam questions is impressive, and possibly unique not just nationally but internationally for an examination at this level.” (for Part II, ie third year)

      “The redoubtable Mathematical Tripos retains justly its reputation for challenging, and sometimes defeating, the best young mathematicians in the country. In doing so, it also offers scope unequalled elsewhere, and in its breadth remains an example to us all.” (Part II)

      “The Mathematical Tripos remains one of the best, by which I mean most difficult, undergraduate mathematics courses anywhere.” (Part II)

      “I suspect that quite a few students who get upper seconds at Cambridge would, for a similar performance, get firsts from some other universities.” (Part II)

      “Part II(B) is comparable to (or even higher than) the standard of courses taught at fourth-year MMath level at other universities” (Part II(B) is the old name for Part II, I believe)

      “The questions set in applied mathematics are similar to those in my institution [Newcastle], but many of these subjects would be taught and examined in the final year rather than the second year. I am therefore satisfied that just over 30% of the students were awarded first class honours.” (Part IB, or second year)

      “Particularly noteworthy is the relatively high percentage of firsts awarded. I am happy that this is a reflection of the high number of very good students in the course who attain the
      required absolute standard of performance…” (Part IB)

      “The level of attainment of students taking the Mathematical Tripos is enviably high, and Cambridge maintains its leading position in the teaching of mathematics in this country.” (Part IA, ie first year – which does in fact have external examiners)

      “The standards achieved by the students were of the highest; I have seen none better.” (Part IA)

  6. Monica Grady Says:

    My problem with widening participation is that the phrase immediately brings to mind the slogan used by Poppleton University: ‘widening participation by lowering standards’

    • Anton Garrett Says:

      That’s unavoidable if you are going to expand higher education to 50% of the population as has been done in the last 20 years.

  7. I don’t see much evidence that widening participation is still a government priority. If anything, policy seems to be directed the other way.

    Foundation programs are excellent for some. We do have one. But you have to decide who it is meant for. It can either aim at students who did not reach the A-level entry requirements because of poor teaching, or it can aim for students who lack the mathematical background, for a variety of reasons. The two don’t mix.

    I have the impression (but no numbers to back this up) that astronomy does better than other areas of physics in attracting students from a wider range of backgrounds

  8. […] * Not by coincidence since all universities are wrapped in admissions right now, one of my favourite bloggers recently discussed similar issues at Sussex. […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: