The evidence against conditional unconditionals doesn’t stack up

Unconditional offers have become the subject of considerable political and media comment in recent years. This article for the first time sets out the evidence of their impact at our university and on our student cohorts.

Much of attention on this issue in recent months has been prompted by the view that making students these kinds of offers is bad for them, both as applicants and later as students. As a university that has been making increasing numbers of such offers, Nottingham Trent has been looking in detail at its data to see to what extent concerns are justified in relation to our approach. In the interests of generating an open and informed discussion, we have decided to put our analysis into the public domain.

Many universities have been making unconditional offers based on factors other than actual academic qualifications for decades. What is new is the use of conditional unconditional offers. These are given to students who are still awaiting the results of qualifications for which they are currently studying and based on their predicted grades.

It is important to be clear at the outset about why NTU makes unconditional offers: to ensure applicants who have the academic pedigree to succeed on our programmes have considered carefully the benefits of making NTU their firm choice; and to enable NTU to attract the most able of these applicants to study at NTU. NTU is the fifth largest recipient of UK applications for undergraduate study and last year recruited more UK undergraduates that any other university. NTU has no need to use conditional unconditional offers to fill its places.

We have been looking at our numbers from two perspectives: what impact conditional unconditional offers appear to have on our applicants; and what impact they seem to make on our students.

The impact on applicants

My university gave a conditional unconditional offer to just under half of its applicants for undergraduate courses in the 2019 cycle. It is worth remembering that offers are only made to those who have applied to us. They have already selected NTU from the vast array of courses and very large range of providers available. If these applicants are predicted, typically, two grades above our published course tariff then they may receive a conditional unconditional offer.

Once offers have been made, an applicant usually accepts one as their firm and one as their insurance choice. Applicants have the whole of the normal UCAS cycle to make up their minds which offer to accept.

One concern expressed around conditional unconditional offers hinges on when a conditional offer made via UCAS is converted to an unconditional offer if the student accepts that provider as their firm choice, in particular if there is an arbitrary time limit. This is the source of the potential, it has been suggested, for universities to engage in “pressure selling”. In our view, there should be no need to require a student to choose before they have all their offers or outside of the standard UCAS decision dates. At NTU we do neither, and nor do we offer any other form of inducement to accept.

Another concern is whether applicants with conditional unconditional offers are more likely to coast through the remainder of their level 3 qualifications and thus miss their predicted grades. National data show it is the applicants with higher predicted grades who miss their grades the most; at NTU it is these very students to whom we make conditional unconditional offers. Our statistical modelling suggests that only a tiny proportion (1.2%) of the variables explaining our students’ propensity to miss their predicted grades can be attributed to holding conditional unconditional offers. This is consistent with the UCAS analysis in its End of Year Cycle report in 2018 which concluded: less than 2 per cent of applicants that missed their predicted A levels by two or more grades in 2018, did so as a result of holding an unconditional firm.

We know from our own surveys that most of the students who accept these offers are positive about them, albeit in only a very small percentage of cases (around 4%) is it the major reason for them choosing NTU. Again this positive sentiment from students is reflected in the national perspective contained in the UCAS 2018 End of Year Cycle report.

The impact on students

OfS has become concerned that students who received conditional unconditional offers may have their chances of succeeding in higher education reduced. In October 2019, it noted that: “we estimate that the impact on continuation rates for entrants in 2015-16 and 2016-17 reduces the continuation rate by 0.65 percentage points once the effect of the other factors has been considered. This means that we estimate around 10 per cent rise in the non-continuation rate, and 185 fewer students who started in these years continued with their studies.”

We have had a very close look at our success rates for holders of conditional unconditional offers. At present, we have two sizeable cohorts of such students to compare with their peers who came in through conditional offers. When considering continuation, we are measuring the number of students who are still enrolled in the subsequent December after their entry i.e. students who had stayed several months into their second year.

For the 2017/18 entry cohort of 18-year-olds, our internal non-continuation rate was the same (8%) for conditional and unconditional offer holders (excluding the unreported external transfers to other HEIs which would result in this figure being closer to 6% for both cohorts). Moreover, and taking account of students’ predicted tariffs and other factors, there is no statistical evidence that the non-continuation rates of our conditional unconditional offer holders differ from what would be expected.

We have also looked at our students’ achievement in their studies, bearing in mind that the 2017/18 intake will not graduate until July 2020. Both groups are identical in the rate at which they fail a module. We also find that, at the aggregate level, conditional unconditional offer holders are doing better when they do pass (Table 1). Statistically controlling for other influential factors (including predicted tariffs, as students coming with conditional unconditional offer have disproportionately higher predicted grades than conditional offer entrants), again we see no evidence of differential attainment between students who arrive at NTU with conditional and unconditional offers.

Table 1: Percentage of year one new entrants achieving grades by degree class

2017/182018/19
ConditionalUnconditionalConditionalUnconditional
First6.6%9.7%6.1%7.7%
Upper Second47.6%48.6%46.1%48.2%
Lower Second39.2%36.5%41.3%38.6%
Third6.6%5.3%6.5%5.5%

Note: the reduction in the % of 1st and 2.1 awarded to both groups in 2018/19 relates to NTU initiatives to reduce grade inflation reported in another Wonkhe blog.

I hope this piece serves to inform the discussion of application processes, and specifically conditional unconditional offers, in the light of the forthcoming publication of the UCAS End of Cycle report for 2019. NTU accepts that politicians, media and public have legitimate interests in the behaviours of universities alongside the legal mandate of OfS to hold us accountable for aspects of our performance. It is our hope that other universities may choose to share similar data in the context that recruitment policies and student outcomes vary across universities.

13 responses to “The evidence against conditional unconditionals doesn’t stack up

  1. This is amusing. If the offers make no perceptible difference to anything at all, why then are they made? If the idea is to firm up students’ commitment to Nottingham Trent, a phone call works wonders. None of this data-thrashing addresses the central question of motivation. It masks it.

  2. I’d argue they’re made because we (in the UK) are at the nadir of 18yo population numbers; the market is competitive and unconditional offers are more likely to have a high conversion rate. The idea is that applicants would rather have an unconditional offer to act as a safety blanket, thus more likely to make an offer their firm or insurance choice.

  3. The VC is more clear on the motivations in his comments in the Times who cover this. We want students who are predicted higher grades than our standard offer to accept us as firm rather than insurance. What we want to show is that, for us, this doesn’t harm the students.

  4. You’ve not included information on the students’ level 3 outcomes – unconditional offers incentivise students to not work as hard. While your own degree level outcomes might be fine, by giving unconditionals you are hurting the success rates of the FECs and 6th forms.

  5. We don’t have a table on the analysis of whether students with CUOs have a higher statistical chance of ‘taking their foot off the gas’ but we have done the analysis (as has UCAS) – hence the conclusion that it’s a tiny proportion of the variables. It’s complex as students with higher predicted grades are more likely to miss them, so we need to control for that (we make CUOs to those students).

  6. Level 3 outcomes are covered in para 9.

    Given that we all have our recruitment targets that must be met, and if unconditional offers were not used, you would presumably find yourself with higher numbers needing to be recruited through Clearing, I wondered if you’ve compared progression between the unconditional and clearing cohorts to establish which shows the most wastage. The presumption is always that Clearing students, coming to terms much later with the institution and effectively with a degree of ‘hobson’s choice’, are more vulnerable to dropping out.

  7. Paul – yes, along with many universities we’ve looked at ‘clearing students’ and their lower persistence rate. But these things are dynamic, and I wonder whether all students who arrived via ‘clearing’ in recent years are less likely to share characteristics.

    Getting students to accept a firm offer early in the process has some obvious advantages for providers, but also for students who can commit to a transition to a particular provider.

  8. Nice to see some actual facts intruding on the HE policy debate as it continues to play out through shouty OfS press releases, stern ministerial pronouncements and broadsheet letters’ pages. Thanks, to you and NTU, for sticking your head above the parapet.

  9. Thank you for sharing this. Is the approach you take to using conditional unconditional offers as set out in this article described clearly for prospective students on your website?

    The article explains that unconditional offers are used for factors other than grade attainment. It also states that students who are predicted typically two grades above the required tariff will typically be those who may receive a conditional unconditional offer.

    Would it be possible to provide a link to where your approach to offer making is explained for potential applicants? It would be helpful to see how this has been done.

  10. Thanks for this piece Mike and thanks to NTU for making this information public.
    Can I just check one thing? As you are using CUCs to get an applicant to make NTU their firm choice does that mean NTU thinks applicants should make the decision to select an institution as the place to study on the basis of their offer status being changed to unconditional?

  11. Dan, no – we’re not trying to ‘make’ people accept our offer. We think that an unconditional offer to students does them no harm, and brings a range of benefits. At the end of the day, especially with auto-release, the choice is absolutely with the applicant.

  12. Dan – applicants will hopefully have a good sense of what a university may do with their application, certainly if there’s proper information on typical offers. An applicant might chose to apply to a university because they know they’ll get a contextual offer, or an unconditional offer. The system is predicated on that predictability, the whole system of typical offers works that way, and unconditional offers do too.

Leave a Reply