Cambridge HSPS Interview Questions for 2027 Entry, With Model Answers

The HSPS Faculty says around 75 per cent of applicants are interviewed each year. The 2025 cycle produced 236 offers on 1,230 applications. This page sets out what the interview actually is for 2027 entry — dated, sourced, College by College — then works through tutor-written questions with model answers that survive the follow-up rather than reciting a framework. The HSPS pack, £180, is where these questions are worked end to end.

What is the HSPS interview for 2027 entry?

Cambridge's main interview period for 2027 entry runs from 7 December to 18 December 2026 — twelve days, counting both ends. Winter Pool interviews follow around mid to late January 2027, and anyone interviewed in the main December window learns the outcome on 27 January 2027. The first of those dates is not negotiable. Cambridge states: "You must keep the interview period free. We cannot reschedule interviews, unless under exceptional circumstances." Not attending ends the application. If there is a school trip, a music exam or a family commitment sitting inside that fortnight, move it in October rather than trying to negotiate it in November.

The second thing to get right is how much interview there actually is. Cambridge's guidance is that "Most applicants will have 1 or 2 interviews lasting a total of 35 minutes to an hour." Read that twice, because most preparation guides — including an earlier version of this page — describe two interviews of twenty to thirty minutes each. That description treats an hour as the floor. Cambridge treats it as the ceiling, and a single interview is a normal outcome rather than a warning sign. The practical consequence is that you are preparing for depth in one conversation, not breadth across several. A rehearsed tour of five thinkers will not fit inside 35 minutes, and trying to make it fit is how candidates end up talking over the follow-up question that was the actual test.

The HSPS Faculty is unusually direct about who gets interviewed: "Everyone with a realistic chance of being offered a place to study at Cambridge is invited to interview – that's around 75 per cent of applicants each year." That one sentence should reshape how you think about the whole application. At many universities the shortlist is the selection and the interview confirms it. At Cambridge HSPS the shortlist is close to the entire field, which means the interview is where the course is decided. The next section puts numbers on how much work that hour is doing.

Format is not yours to choose either. The Faculty says: "Most interviews take place online with some exceptions. Each College organises its own interviews so each format may differ. Expect to meet interviewers from more than one subject offered within HSPS." Two things follow from that. First, the College you name on your UCAS form — or the College that picks you up in the Winter Pool — decides whether you sit in a room in Cambridge or at your own kitchen table, and you will not get a say. Second, you will almost certainly face an academic from a discipline you did not lead with. A candidate who has read only politics will meet a social anthropologist, and the anthropologist will not pretend otherwise for the sake of a comfortable conversation.

How competitive is HSPS, and where is it actually decided?

No competitor page ranking for this query publishes an applications-to-offers figure for HSPS, which is odd, because the numbers are published and they change how you should prepare. Here are the two most recent complete cycles side by side.

Start with the ratio. In the 2025 cycle, 1,230 applications produced 173 acceptances, which is 7.1 applications for every place taken up, dividing 1,230 by 173. Cambridge says of itself: "On average across all subjects, we receive about six applications per place." So HSPS runs above the University-wide average, though not dramatically so. It is harder than typical Cambridge, not exceptionally harder, and anyone telling you it is the back door into the University has not looked.

Now combine that with the interview rate, which is where the picture actually gets interesting. Around 75 per cent of 1,230 applicants is roughly 920 people interviewed. Set 920 against 236 offers and you get about 3.9 interviewees per offer. Set it against the 173 places actually taken up and you get about 5.3 interviewees per place. In other words, roughly four in five applicants clear the paper stage, and then four out of five of those are turned down after the interview. Almost the entire selection happens in 35 to 60 minutes of conversation. That is an unusual shape for a UK admissions process and it justifies putting the bulk of your preparation time into being interviewed rather than into polishing the form.

One more piece of arithmetic worth knowing: 63 of the 236 offers did not end in a place taken up, since 236 minus 173 is 63. Some of those candidates missed their grades and some chose elsewhere. The published statistics do not distinguish between the two, and neither should you when you are quoting them.

Comparing cycles is where most guides start inventing a trend. Applications fell by 29, from 1,259 to 1,230. Offers rose by 8, from 228 to 236. The success rate moved from 13.8% to 14.1%, a rise of 0.3 percentage points. Nothing in that says HSPS is getting easier. It says one cycle had slightly fewer applicants and slightly more offers, which is what two adjacent years usually look like in any subject. Two data points are not a direction of travel, and saying so out loud is exactly the habit HSPS interviews reward.

Two features of the intake stand out. Maintained-sector candidates made 73.9% of Home UK-school applications but took 83.4% of offers and 79.6% of acceptances, so the share rises at every stage rather than falling. And the gender skew is among the widest of any Cambridge social science course: 71.0% of applications and 69.2% of acceptances were women.

Treat both rows with care, though, because they come from different tables with different denominators. The sector figures cover Home UK-school applicants only, while the gender figures cover everyone, and the gender rows do not sum to the headline totals: 348 men plus 851 women is 1,199 against 1,230 applications, and 52 plus 117 is 169 against 173 acceptances. A small number of records sit outside the split. That is ordinary in published admissions data, and noticing it is worth practising, because spotting exactly that kind of gap is what an interviewer is testing when they slide a table across the desk and ask what it shows.

Which Colleges set an HSPS assessment, and when do you sit it?

This is the fact the search results get wrong. At least two pages currently ranking for HSPS interview queries still headline a pre-interview HSPS admissions assessment. There is no such thing for 2027 entry. Cambridge's admission-tests page lists the course as follows: "Human, Social, and Political Sciences (HSPS) | College admission assessment You don't need to register in advance." If you have been told to register for an HSPS paper, you have been told something that is no longer true, and you should not spend your October on it.

The assessment that does exist sits after shortlisting and is set by three Colleges only. The University names them: "Human, Social and Political Sciences Colleges that ask you to take an assessment Hughes Hall, King's and Newnham. Assessment format Details will be provided by the relevant College." Pair that with each College's published interview format and the picture is unusually concrete for something most guides describe in the abstract.

Look at what that table means in practice. Your College choice determines two separate things at once: whether you sit an additional written assessment at all, and whether you are interviewed at a desk in Cambridge or on a screen at home. Two of the three assessment Colleges interview everyone online; the third interviews everyone in person. Nothing about that is negotiable after you submit the form, and the University does not publish the assessment format in advance, only that the relevant College will send you the details.

It is worth understanding how different this is from the courses that do run registered papers. A friend applying for Law, Economics or Engineering registers months ahead and sits a fixed paper on a fixed date; our guide to university admissions tests covers what those involve. HSPS applicants register for nothing. The only entry in the diary is the interview window, and for three Colleges an assessment that will be arranged around it.

So what should you actually do? If you applied to Hughes Hall, King's or Newnham, wait for the College's own instructions and read them the day they arrive, because the format is theirs to define and general advice about admissions tests will not describe it. If you applied anywhere else, and are not pooled to one of the three, the interview and your two pieces of written work are the entire assessment. That is a short list of things to be good at, which is either reassuring or alarming depending on how much attention you have paid to the essays.

Why your two pieces of written work drive the conversation

The Faculty requirement is simple: "HSPS applicants are required to submit 2 pieces of written work, such as an essay, prior to interview." What almost nobody builds preparation around is the sentence that follows on the Faculty's interview page: "Your 2 pieces of submitted written work will most likely inform the interview discussion."

That is the most under-used sentence in HSPS preparation. It tells you, in advance, roughly where a good portion of the questions will come from — and unlike the rest of the interview, it is a source you control completely. You choose which essays go in. Choosing the piece that got the best mark is the obvious move; choosing the piece whose argument you can still defend eight weeks later is the better one. A polished essay on a topic you have half-forgotten is a trap you set for yourself in October and spring in December.

The drill is straightforward and most candidates skip it. For each of the two pieces, write out three things on a single sheet: the claim the essay actually makes, once you strip the introduction; the strongest counterexample or counterargument you know of, stated in its best form rather than a straw version; and what evidence would make you change your mind. Then read the essay again looking for the weakest join in the argument — the sentence where you moved from one claim to another slightly larger one without earning it. That join is where an interviewer will press, because it is where pressing produces thinking rather than recall.

There is a second controllable input. My Cambridge Application is where you record what you have actually studied, and Cambridge is explicit about why that matters: "For example, details of topics you've covered as part of your A Level courses (or equivalent) may help our interviewers to decide which questions to ask." The form is, in effect, a menu you hand the interviewer. Fill it in carelessly and you invite questions on the module you rushed; fill it in accurately and the questions land on ground you have covered. If a topic is listed there, treat it as examinable, and if your A-level tuition covered a unit in more depth than the syllabus required, that is worth recording rather than leaving out for modesty's sake.

Politics questions with model answers that survive the follow-up

The strongest HSPS answers are not the ones with the most references. They are the ones that turn a vague question into a claim you could be wrong about, and then handle being pushed on it. Two questions, each with a weak version and a strong one.

Question: is liberal democracy in crisis? A weak answer starts listing: populism, polarisation, misinformation, a couple of election results. It is not wrong, it is unfalsifiable. Nothing in it can be checked, so there is nothing for the interviewer to push on except to ask what you mean by crisis — which was the question you should have answered first.

A strong answer does the definitional work out loud and then commits to a measure. "It depends what would count as a crisis, and I think there are two different claims here. If we mean falling participation, there is evidence: turnout at the 2024 UK general election was 59.7%, down from 67.3% in 2019, a fall of 7.6 percentage points in five years. If we mean a representation gap, that is a different claim with different evidence: Reform UK stood 609 candidates, took 4,117,610 votes and 14.29% of the UK vote, and won five seats — 823,522 votes per seat, dividing 4,117,610 by five. I think only the second is a crisis of the system rather than of the parties, and I am not sure the first is a crisis at all."

The follow-up is where the answer is actually scored, and it will come. An interviewer will not accept the figures and move on; they will ask whether a disproportional result is a failure of the system or the system doing precisely what it was designed to do. A first-past-the-post election is not attempting to deliver proportionality, and a candidate who has read anything on majoritarian versus consensus models of democracy should say so rather than defend the implication that disproportionality is by itself illegitimate. The strong move is to separate two claims: that the system is behaving as designed, and that the people living under it still consent to the design. The seat count speaks to the first. The turnout figure speaks, weakly, to the second.

Question: should voting be compulsory? The trap is to argue from the turnout figure straight to the policy. A 7.6-point fall tells you how many people voted. It does not tell you who stopped voting, or why, and compulsion would fix a number without necessarily fixing whatever the number is standing in for. A strong answer says that plainly: "Compulsory voting raises turnout by construction, so the interesting question is whether it raises legitimacy. That depends on whether current non-voting signals alienation or contentment, and an aggregate turnout figure cannot distinguish those two." Then name the evidence you would want — who the non-voters are by age, income and constituency — and admit you do not have it. Interviewers consistently reward a candidate who identifies the missing data over one who guesses at it confidently.

None of this is really about political knowledge. The interviewer is checking whether you can take an evaluative word — crisis, legitimate, fair, representative — and convert it into something with a truth value. The Part I politics papers, whether POL1 The Modern State and its Alternatives or POL2 International Conflict, Order, and Justice, are built on that single move, which is why the interview tests it before you have read a word of either.

Sociology and social anthropology questions with model answers

Question: is poverty best understood as a lack of income? A weak answer reaches for the headline number and stops there. The Joseph Rowntree Foundation reports that more than one in five people in the UK, 21%, were in poverty in 2023/24 — 14.2 million people, of whom 7.9 million were working-age adults and 4.5 million were children. That leaves 1.8 million who are neither, since 14.2 minus 7.9 minus 4.5 is 1.8. Every word of that is accurate, and all of it is recitation. It answers a different question: how much poverty is there, rather than what poverty is.

A strong answer uses depth rather than headcount, because depth is where the measurement argument lives. Of those 14.2 million, 6.8 million were in very deep poverty, with incomes at most two thirds of the poverty line — almost half of everyone in poverty, since 6.8 divided by 14.2 is 47.9%. The Foundation records that as both the highest absolute number and the highest proportion on record, going back to 1994/95. Now the answer has somewhere to go: a headline count can sit almost still while the experience underneath it gets sharply worse, so a single line-crossing measure is not describing the same phenomenon from one decade to the next. Income is the instrument, not the thing being measured.

Expect the interviewer to attack the measure rather than the number. If the poverty line is set relative to median income, can poverty fall during a recession simply because the median falls? The honest answer is yes, and that this is an artefact of the metric rather than an improvement in anyone's life. That is an argument for reading the relative measure alongside a depth measure and a material deprivation measure, not for abandoning it. Saying "this measure has a known failure mode and here is what I would read next to it" is a far stronger position than defending a measure as though it were the phenomenon itself.

The social anthropology move on the same question is different again, and worth having ready, because Social Anthropology: The Comparative Perspective (SAN1) is one of the Part I papers you are most likely to be interviewed on. An anthropologist would ask what counts as deprivation to the people being counted. A poverty line is an artefact of a particular state's administrative capacity and its statistical conventions; the ethnographic question is what a decent life is taken to be locally, and whether the line tracks it. That is not a rejection of the statistics. It is a claim about what statistics of that kind can and cannot see, and it is defensible for ten minutes under pressure, which is the real test.

The sociology paper, Introduction to Sociology: Modern Societies I (SOC1), rewards the same discipline in a different register: name the measure, name what it misses, and say what you would look at next. If you can do that with poverty, you can do it with social mobility, with educational attainment, with crime statistics, and with whatever unfamiliar dataset the interviewer produces. The topic is interchangeable. The habit is not.

Interdisciplinary questions, and what HSPS Part I actually contains

Before the interdisciplinary questions, a correction this page itself needed. An earlier version described HSPS Part I as four component disciplines, "Politics, Sociology, Social Anthropology, and either Criminology or Psychology". That is wrong, and it is wrong in the same way across a great deal of the internet, including pages that rank well for HSPS queries.

The Faculty sets out the real structure: "In Part I, students take four introductory papers, students should select three papers from: One from: The Modern State and its Alternatives (POL1) or International Conflict, Order, and Justice (POL2) Social Anthropology: The Comparative Perspective (SAN1) Introduction to Sociology: Modern Societies I (SOC1)". The fourth paper comes from a separate list that includes Psychology (PBS02) along with archaeology and biological anthropology options. Psychology is genuinely available in the first year. Criminology is not.

Criminology does exist in HSPS, but later and only as half of a pair. The Faculty lists three single-track options and five joint-track options for Part II, and Criminology appears in exactly one of them, Sociology and Criminology, chosen towards the end of the first year. If you tell an interviewer that you applied to HSPS in order to study Criminology, you have told them you did not read the course structure, and you will have done it in the first two minutes. Fixing that is free.

Question: why do people obey the state? This is the shape of question HSPS uses to see whether you can hold more than one lens. A politics answer reaches for authority, legitimacy and the conditions under which coercion is accepted as rightful. A sociology answer reaches for norms, socialisation and the institutions that make obedience unremarkable rather than chosen. A social anthropologist asks what obedience looks like in settings where the state is thin on the ground, and whether the question has quietly assumed that the state is the relevant authority at all.

The HSPS answer is not all three at once, which sounds like indecision dressed as breadth. It is holding two of them, saying which one your evidence actually supports, and being explicit about what would make you switch. "I think the sociological account explains routine compliance better, because most obedience is not deliberated at all. But it struggles with the cases that interest me most, which are the ones where people stop obeying, and there the political account of legitimacy does more work." That is a position, it invites a follow-up, and it survives one.

This is also why the Faculty warns you to expect interviewers from more than one HSPS subject. If you are rehearsing with a tutor, ask to be interviewed from the discipline that is not your strongest; our subject specialists cover politics, sociology and psychology separately, and the friction between how they each read the same question is the point of the exercise, not a scheduling inconvenience.

A Worked Question in the Pack's Own Shape: A Table That Argues With Itself

The HSPS pack runs its ten questions in three separate layers — the bare question on its own, then Prompts, then a full worked answer — rather than the essay-length answers above. Here is one built the same way, on a topic the pack itself does not cover, so nothing below is something you would otherwise pay for.

Question

Imagine a university with two departments, Department A and Department B. In Department A, 90 men apply and 30 are admitted; 10 women apply and 4 are admitted. In Department B, 10 men apply and 1 is admitted; 90 women apply and 10 are admitted. In both departments, taken separately, women are admitted at a higher rate than men. Work out the combined admission rate for men and for women across both departments together. Then explain what you find, and say whether it shows discrimination.

Prompts

• Before combining anything, write out four numbers for each department separately: applicants and admissions, for men and for women.
• Once you have both departments' rates, ask what you are actually averaging when you add them together — two groups the same size, or two very different ones?
• Look at where each gender's applications are concentrated. Does that concentration do any work here?
• A gap between two combined rates does not by itself tell you where in the process it arose. What further evidence would locate it?

Suggested answer

Where the reasoning goes wrong. A candidate who jumps straight to the combined figures sees that men end up admitted at a noticeably higher overall rate than women and concludes that something in the selection process must be favouring male applicants — perhaps in how references are weighed, or how borderline cases are judged. That conclusion treats the two headline percentages as if they were simply comparable, two numbers measuring the same thing. They are not: each is an average taken over a different mix of Department A and Department B, and the question never said the mix was the same for both groups. Reasoning from the total straight to a claim about how individual decisions were made skips the one step that mattered.

Where it holds up. Write out both departments first. Department A: men 30 admitted out of 90 applied, which is one third, about 33%; women 4 out of 10, which is 40%. Department B: men 1 out of 10, which is 10%; women 10 out of 90, which is about 11%. So in Department A, women's admission rate is 40% against men's 33%, and in Department B it is 11% against 10%. Women are ahead in both, exactly as given.

Now combine. Men: 90 plus 10 applied is 100; 30 plus 1 admitted is 31; that is 31%. Women: 10 plus 90 applied is 100; 4 plus 10 admitted is 14; that is 14%. Men's combined rate is more than double women's, despite trailing women in every department taken on its own.

The reversal has a mechanical cause, not a hidden one. Ninety of the men's hundred applications sit in Department A, the department admitting around a third of everyone. Ninety of the women's hundred applications sit in Department B, the department admitting roughly a tenth of everyone. A combined rate is a weighted average, and each group's weight falls almost entirely on the department it applied to in bulk. Men's total is pulled toward Department A's generous rate; women's total is pulled toward Department B's tight one. No single admissions decision needs to have treated a man and a woman differently for the totals to move apart this far. The arithmetic alone produces the gap once the two groups apply in such different proportions to the two departments.

That does not settle the discrimination question; it relocates it. The combined figures cannot tell you whether the process itself is fair, because they are consistent with a process that judges every applicant identically. What the reversal does tell you is that 90% of the women in this example applied to the harder department and 90% of the men applied to the easier one, and that pattern is what wants explaining — a question about why applicants sort themselves so unevenly between the two departments in the first place, which a table of admission rates cannot answer on its own. Anyone who says the aggregate figure alone proves bias, in either direction, has stopped one step before the numbers actually run out.

The examiner is not testing whether you have met Simpson's paradox before. It is testing whether a single combined number stops you short of a conclusion the table cannot support — the same discipline this page has already used on the real Cambridge admissions data above, where the sector and gender rows do not sum to the headline totals either.

Grades, subjects and the essay-subject condition nobody mentions

The course page and the Faculty page say subtly different things about subject requirements, and the difference matters a great deal if you are choosing A-levels now. The line that gets quoted everywhere is that HSPS requires no specific subjects. The full sentence from the Faculty is: "There are no required subjects for entry to HSPS. Colleges do normally require A*/7 in an essay-based subject." Both halves are true. Only the first half travels.

Read the last two rows carefully, because they are easy to misuse. The 69% figure describes what entrants achieved — Cambridge's own wording is that most HSPS students who studied A levels and started in 2018, 2019 and 2023 "achieved at least A*A*A (69% of entrants)". It is an outcome, not a condition. It does not mean Colleges make A*A*A offers, and quoting it as though it were the offer level would be exactly the sort of slide from one claim to a larger one that this page keeps warning about. What it does tell you is that the published A*AA minimum is the floor of the room, not the middle of it.

If you are still choosing subjects at sixth form entry, the practical reading is generous rather than restrictive: almost any combination is admissible, provided at least one subject requires you to build a sustained written argument and you are on course for the top grade in it. Neither the University nor the Faculty sets GCSE requirements for HSPS. That is not the same as GCSE grades being invisible on a UCAS form, but it does mean there is no threshold to clear and no reason to talk yourself out of applying on the basis of one bad result two years ago.

What Cambridge now prohibits in the interview room

This one is absent from every competing guide, and it is new enough that most candidates have not thought about it. Cambridge's guidance for 2027 entry says: "Any use of AI tools during an interview hinders this approach and is therefore prohibited. Various methods will be used by interviewers to randomly check that AI tools are not being used."

Because most HSPS interviews take place online, this rule bites harder here than it does on courses interviewed in person. An online interview is conducted on a device that could, in principle, be running something else. Cambridge has clearly thought about that, and the reference to random checks means the burden is not on them to suspect you first. Close every other window. Put the phone in another room rather than face-down on the desk. If a second screen is visible in frame, move it.

There is a deeper reason to take this seriously than the rule itself. The pause after a hard question is not dead air to be filled; it is the part of the interview the interviewer is watching most closely, because it is the only unrehearsed thing in the whole conversation. Thinking visibly — "I want to separate two things in that question before I answer" — is worth more than a fluent answer arriving instantly, and anything that shortens the pause makes you look less like a candidate who can think and more like one who can retrieve.

One further rule of the day belongs alongside it. If you are ill or something goes wrong, tell the College before the interview, not afterwards: "Any illness, injury or personal circumstances that you've reported after your interview won't usually be taken into account when the College considers your interview performance." Combined with the no-rescheduling rule, that means an uncomfortable decision has to be made on the morning rather than in hindsight. Have the College admissions office contact details somewhere you can find them in thirty seconds during that fortnight.

The 2027-entry timeline, start to decision

Everything above hangs off four dates. HSPS is applied for through UCAS under course code L000, and Cambridge notes that its deadline of 15 October "is earlier than most universities" — which catches out candidates whose schools work to the January cycle.

Working backwards from that table gives you a preparation plan rather than a reading list. The two pieces of written work are chosen and submitted well before December, so they are an October decision, not a December one. Mock interviews are worth doing in the last three weeks of November, close enough that the habits stick. And if you are pooled, the January window means the work does not stop on 18 December, which is worth telling a candidate in advance so that a Winter Pool letter reads as a second chance rather than a rejection.

Frequently Asked Questions