UCAS Personal Statements and AI

[Ceramiq]

I don't really understand how universities are going to be able to use Personal Statements in future: it is so easy to write and/or improve a fantastic PS in minutes using Claude. Any thoughts?

A different question from false accusation but interesting.

At UCAS the issue is similarity detection and I do think any particular LLM used at scale on such a limited topic as the PS may leave patterns, all the more so when the PS becomes segmented.

I see some advantages in dispensing with the PS but grade inflation has got in the way. Norm referenced A level grades would make this much more feasible and TBH I think it no bad thing for applicants to know their rough centile standing against their peers. Not to make a meal of it, just to act as a reality check.

When accomplishment is high one doesn’t want overly tight grade bands, or to discourage candidates, but attainment at university doesn’t suggest that this is the real story.

It is tricky.

You stated ‘legitimate interest and public task’ but didn’t say what the task and legitimate interest were in holding onto data they do not use.

Interesting. I was at a ucas conference last year which included a panel of admissions tutors from St Andrews, Oxford, Cambridge and Imperial. They all said that every is PS read. It was an international conference though, not domestic, don’t know if that made a difference. We also got an hour-long lecture from UCAS on the PS, they didn’t imply it was going anywhere - quite the contrary tbh 🤷‍♀️

@titchy@clary@Kiminki
I think students should have the option of writing a different personal statement for different universities.
Based on the student room- many students who are applying for Economics at the majority of universities, can't apply to LSE for Economics because it is one of the only universities that requires Further Maths. LSE then rejects their application for similar courses like Accounting and Finance and Management because they have written an Economics focused ps for other universities but still want to try and get into LSE. This is despite Accounting and Finance and Management at LSE having many Economics modules.
Many students apply to Oxford for Economics and Management but not all universities offer this course. Then they are disadvantaged if they apply for Economics only or Management only at universities like LSE and Bath who do not offer Economics and Management. LSE and Bath say they do want the personal statement tailored to their course so are rejecting students who have not written a focused Economics or Management statement.
This is really difficult for students to navigate-when they are just trying their best to get into the best universities for going into a finance-related career. Apparently, the target universities for finance are Oxford, Cambridge, LSE, UCL and Warwick but they don't offer the same courses.

My dd applied for two fairly unrelated courses (English lit and Class civ) with one PS that covered both.

I suggest if a YP is applying to a uni which uses the PS (only a few do, we have established and it’s easy to ask) the you tailor it to that uni. There is a lot of overlap on the courses you mention @fairyring25. A student without FM will struggle to get on an econ course on the unis you mention tbh.

But yes the option of multiple PSs might be useful. Only for very specific cases tho surely.

I think the problem is not necessarily in the use of AI as some posters have described above - as an editor or critical friends, but in the fact that many students use AI badly.

I work in a university and we ask for personal statements for a specific opportunity we offer. Each year we receive many statements that are wholly or partially AI, very obviously. Those applications are disregarded because we prohibit use of AI in our guidelines.

This becomes a digital equity issue because it is our most disadvantaged students who are not being taught the ‘right’ ways to use AI in schools.

English Literature and Classical Civilization are less competitive (ie fewer applicants per place, not less academically challenging) than Economics and related courses. It is easier to imagine admissions assessors needing to score Economics PSs much more harshly than, say Classical Civilization PSs, just to reduce the sheer volume of applicants. Hence any irrelevant information because a PS is covering more than one course being more risky the more competitive the course.

Maybe A-levels need to be graded on a 1 to 9 scale with half-points? This seems to be a popular grading scale among standardized tests (IELTS, TMUA, ESAT, TARA) that claim to offer reliable discrimination between candidates. I am very set against precise percentages just because we all know that grading is an imperfect science that causes enough stress as it is.

Exam grades are very easy for universities to use but are they really the best way of deciding the best candidate for a course? Above a certain level that ensures they can cope with the rigour of the course, would it not be better if there were some assessment of level of interest? But how can this be measured?

There are definitely arguments for a PS that judges applicants on their understanding of the course but the single PS format isn't good at that.

I don’t think it’s going anywhere, either. Nor do I think we will be seeing norm referenced exam grades.

This is the function of interview, IMO.

Edit : A function of interview.

This would help, but the fundamental problem is grade creep.

Your point about digital equity is very important.

Interviews are very expensive for universities, even with zoom. And are themselves not a particularly good way of picking candidates as too prone to bias.

By grade creep, do you mean that grades increase marginally over time for a given calibre of student? And to what do you ascribe this? Examinations with very explicit mark schemes (as GCSE and A-level have) where students are taught precisely what to do in order to gain points on exams tend to optimize increasingly for grades and push them up over time without the students truly becoming better educated in any meaningful way. But what do you do to prevent this if mark schemes are extremely explicit?

The reality is that most courses at most universities will take/make offers to anyone who meets their minimum grade requirements. In fact many will take students who don’t make their grade requirements. These courses don’t care what you write in your PS, they want bums on seats or £ in the bank. (When unis write to prospective students referencing their PS to say how much they hope they will accept their offer they’re in marketing mode, it’s not because your DC wrote the most amazing PS. Yes, they’ve read it, but they’ll be sending that letter to everyone, it’s not a value judgement. Sorry but that’s the reality.)

Even at Oxbridge, for science courses, they don’t care about PS. They look at grades, admission tests and performance in interviews - which are entirely problem based.(They will say they look at PS but as long as you can write a few sentences about a book/course/competition/project you did, you will be fine, they are not making final offer decisions on the basis of PS.)

It’s a farce which universities participate in because they don’t want to admit the true situation: in the main they are desperate for students.

A few massively over subscribed courses do have to make selection decisions on the basis of the info in front of them and maybe for some that will include the PS, but a PS will not trump the highest predicted grades or some contextual info. (A further extension of the farce is unis - I’m looking at you LSE - who explicitly tell students that they didn’t get offers because on their PS.)

I know an admissions tutor at a top uni where they take students who have missed their offer rather than go into clearing. At that point they pick the students who will tick the most boxes on their widening participation criteria. They don't look at the PS - despite what they say at open days.

The amount of time, resources, webinars, open day presentations that are spent on the PS are completely disproportionate to its value. Students would spend their time more wisely improving their A level grades. Universities should step up and be honest about this.

AI just makes this all the more obvious.

Yes, on a few courses, at a few universities where there are interviews, the PS may be a useful ice breaker, but to really test students interviewers need to get candidates off their prepared material and into the unknown and unfamiliar, so even there, the value is marginal.

[The only exceptions are highly vocational courses like medicine where they want to see specific experience.]

I don’t think the problem is the mark scheme, it’s the extremely prescriptive syllabus. It’s the tight syllabus which means the range of possible exam questions is very narrow which means a student can effectively prepare by rote learning the mark schemes from previous exams.

The only exceptions are highly vocational courses like medicine where they want to see specific experience.

Very few medical schools look at PS. It would simply take too long. Interview offers are based on grades and UCAT - mostly UCAT. Once you’ve taken UCAT where you apply is pretty much purely based on your UCAT score with medical schools stratified by its score. There are a couple of exceptions that look at voluntary work etc but they send out extra forms requesting this information rather than relying on the PS. Medical schools might reference the PS in the interview but the scoring is based on the response to the interview question not what was in the PS.

I am aware of this as a former RG admissions tutor. A lot is being done to address bias and I think this is improving.

The tight syllabus and precision mark schemes are to some extent part and parcel of the same concept and lead together to examination optimization practices. TBH I had that feeling at university and my DC who is assessed 75% on coursework and 25% on examinations says that examinations are a ridiculously poor way of assessing what a student is capable of versus coursework.

@Ceramiq I also think that grade creep is to do with the mark schemes. The teachers teach to the test. The students don't need to think in the exam, they just regurgitate what they have learnt.
@WW3 I think since 2015 the A-level specifications have more content in general. Theoretically, this would allow the exam boards to ask many different types of questions but they don't. They use the same types of questions with similar number of marks every year. Therefore, if the students have an experienced teacher, each question is broken down for them so they know exactly what to write for each question or how to solve the problem presented. It also means you can't distinguish between those who rote-learn information/have a good teacher and those who genuinely have deeper critical thinking or problem-solving skills.

Though course work is vulnerable to use of AI - not necessarily to write it directly but to neatly present all the information required to write it without any particular skill or knowledge.

In terms of universities just wanting to attract numbers, the situation is different for Scottish pupils in Scotland. The number of places are capped and universities have no incentive to lobby to allow more places as fees (£1820, paid for for most students) and money from government don’t cover costs. Hence universities may have spaces for students from the rest of the UK (at UK fee levels) but not Scottish students. Political pressure means that their over subscription criteria is mostly prioritising contextual students. This is principally measured by post code and thus housing density which favours SNP supporting areas. A couple of years ago to went to the extent of no non-contextual students even being considered for nine courses at Edinburgh University, including Law and Economics.