Undergraduate Admissions Handbook 2012-13

1.4 Interviewing and selection

1.4.1 Selection criteria

How do we assess applicants for Cambridge?
Admission to Cambridge is highly competitive and selectors are required to make very difficult decisions between applicants, based on each applicant's potential to benefit from and flourish in studying the Cambridge course for which they have applied. Applicants are assessed individually using all the information available to the selectors. Decisions about applicants are made on the basis of the following:

• academic record, including GCSE (or equivalent) grades, AS Level (or equivalent) marks and grades, and A Level (or equivalent) marks/grades or predictions
• school/college reference
• personal statement
• relevant contextual data
• submitted work, where requested
• test results, where a written test forms part of the assessment
• performance at interview, if interviewed.

All of these indicators are considered when we make admissions decisions, and all are important.

What are we looking for?

• Academic ability: Selection is based overwhelmingly on academic potential. Successful applicants are amongst the most able students in their school/college in the subjects most closely related to their chosen university course. They are normally predicted high grades in A Level subjects (or equivalent) appropriate for their chosen course. They show intellectual flexibility, analytic ability, clarity of thought, and an ability to argue logically. They are not afraid of complex and challenging new ideas.
• Motivation and suitability for the chosen course: Successful applicants have a strongly demonstrated commitment to their chosen course. This might be shown by, for example, an exploration of the subject beyond the bounds of the school curriculum through wider reading, or by appropriate work experience or non-school activities.
• Commitment and self-discipline: Successful applicants are self-motivated, self-disciplined and academically committed. This can be shown by good time management in the balance of their various academic and personal commitments.

What are we assessing in our written tests and interviews?

• Interviews may be used to explore any questions that may be outstanding from the written application.
• The main focus of interviews and written tests is to explore academic potential, motivation and suitability in the chosen field. This may include problem-solving ability, the assimilation of new information and ideas, intellectual flexibility and analytical reasoning. How these qualities are assessed and the relative weight placed on each will inevitably vary from subject to subject.

Mature students
The same criteria are used for the selection of mature students. Academic potential is judged by evidence of recent academic work at a level sufficient to show that the applicant can cope well with the rigorous Cambridge course at the point of entry.

1.4.2 Interviews

Information to be sent to applicants attending interviews
Detailed interview arrangements will vary between Colleges and between subjects within an individual College. It is thus important that clear information is sent to each applicant in sufficient time to enable travel arrangements to be made and any pre-interview material to be prepared. Such information should include:

• date, time, duration and location of all interviews and details of where to report to
• names of interviewers and whether the interview is a subject interview or a more general academic interview (The Admissions Forum has agreed that the term 'tutorial interview' should not be used.)
• details of any written work to be sent
• details of any written tests
• information about interviews in other Colleges or within a faculty/department (if any) and where to report to
• arrangements for staying overnight if coming from a distance, or required to do so by the interview schedule, and where to report to
• domestic information about meal times etc
• map of how to reach the College.

Advice is given on the International Students Team website regarding visa requirements for interview.

The Admissions Forum has agreed that applicants should not be charged for overnight accommodation or essential meals.

Disabled applicants, those with health problems or those requiring additional support should be given clear instructions as to their interview arrangements, and special arrangements may need to be made by both staff in the College and those interviewing the applicant to ensure that the interviews comply with the agreed equal opportunities and disability guidelines. For further information see Section 2.7.

Information to be sent to applicants not being called for interview
Applicants who are not being called for interview should be informed by letter. The letter should include information about the timing of any decision on their application. If the applicant is not being invited to interview because it is clear that, on the basis of the information available, their application will be unsuccessful, then it is acceptable, indeed recommended, that this is made clear in the letter informing the applicant that they are not being invited for interview. It is unkind to encourage false hope in such applicants by delaying notification of the decision until January.

Timing and conduct of admissions interviews
Interview dates for some small subjects have been agreed by the Admissions Forum and are shown in the table below:

* The Directors of Studies and Colleges have not made any formal agreement to confine interviewing to these days, and individual Colleges may well choose to interview earlier or later, in the light of the commitments of their Tutors etc.

All applicants invited for interview in Cambridge should be seen by at least two interviewers at their preference College (or, in the case of open applicants, at their allocated College). The total duration of the interview(s) for a given applicant should not normally be less than 30 minutes.

Interviewers should be selected to ensure that, as far as is possible within the availability of trained and experienced people, those used reflect a balance of academic interests, gender, ethnicity and educational experience.

Colleges conducting second interviews as part of an agreed subject arrangement, or Colleges interviewing applicants from the Winter Pool, should not normally arrange more than two interviews for each applicant.

Colleges should show respect for religious holidays and observances when arranging interview dates and the timing of interviews throughout the day.

Detailed guidelines for the conduct of admissions interviews can be found in the ‘Information for Admissions Interviewers for the Cambridge Colleges’ (see Appendix A) which is sent to all interviewers.

Monitoring information about recruitment

Should a College wish to collect feedback, they are reminded that such questions are inappropriate for an admissions interview, and should be included in a separate questionnaire which can, if the applicant so wishes, be returned anonymously in a separate envelope to the College admissions office. This questionnaire should NOT be available to interviewers and should include a statement such as:

‘We welcome some feedback on our prospectus and open days and would be grateful if you could complete this questionnaire. Please note, however, that what you say will not be used in the selection process and is for monitoring purposes only.’

1.4.3 Assessment of GCSE scores

The Admissions Forum has for some time been considering how the assessment of the academic results of each applicant can be made more objective, in the interests of fairness and transparency in the admissions process. The method recommended below is the outcome of research undertaken by the Admissions & Data Services team in CAO at the direction of the Admissions Forum's Working on Admissions Research and was agreed at its meeting in November 2010.

The scoring system is restricted to applicants educated in English schools and colleges, since it has not to date been possible to obtain equivalent information for schools outside England.

GCSE scores
Count the number of A* and A grades (NA* and NA) and ignore all other grades. Score 1.0 for each A*. Assign a value F of between 0 and 1 to each A grade, where F depends on the school/college capped GCSE score S (see below) as follows:

Represented graphically the relationship between F and S is:

Thus an applicant's GCSE score G is:

It has been agreed that all GCSEs taken should be counted towards this score, i.e. there is no cap, but short-course GCSEs should be excluded.

School/college score at GCSE
The best way of assessing the academic achievement of a school at GCSE is to look at the average (capped) GCSE point score. Up until 2003, this was the headline measure of school performance used by DfES (then DCSF, now DfE). Since then, the methodology used in the official Performance Tables has been the subject of constant change and since 2004 has conflated GCSE results and 'equivalent' vocational qualifications. Rather than relying on official measures of questionable value, Cambridge Assessment calculate for us the average (capped) GCSE point score for every English school using the 2003 methodology but the most recent available GCSE cohort data. This measure is included in the Schools Database produced by CAO.

This point score is calculated for each student as follows:

Thus 8A* = 64 points.

Only the best eight results achieved by each student contribute to the score. The average point score for the school is then the average of the individual scores achieved by all students who took GCSEs.

The average point score ranges from 20 (very poor) to 61. In general, for the schools that produce Cambridge applicants, those scoring:

• 60+ are the very top independent and selective state schools
• 50-60 are the good independent and selective state schools
• 40-50 are good comprehensive and other non-selective state schools and weaker independent schools
• 30-40 are average non-selective state schools
• below 30 are poor academic schools with probably few high achievers in each class; applicants from such schools may well be assessed through the Extenuating Circumstances Form (Section 2.1)

The advantages of this simple score are that it is not based on the type of school – there are many state schools that score more highly than most independent schools – and it provides a proxy for the sort of academic support that is likely to have been given to the applicant: a proxy that when used in the GCSE scoring formula detailed above has been shown to improve the correlation between this measure and Tripos performance here.

School/college score at A Level

The DfE/DCSF website also lists for 2007 onwards the average A Level point score for each school/college, based on a scoring system that assigns points as follows:

and half these scores for an AS Level.

In all cases two average scores are reported: the average point score per examination entry, and the average point score per student. Both are useful for the purposes of assessing AS Level results and A Level predictions. The former has the advantage of naturally being capped.

1.4.4 Contextual data

Data provided by UCAS

From the 2011/12 admissions round UCAS will be providing the following contextual data for each applicant:

• Percentage achieving 5 A*-C GCSEs (including English/Welsh and Mathematics)
• Average capped GCSE points score
• Average QCA points per qualification per entry
• Average QCA points per qualification per student
• Percentage of students entitled to free school meals
• Percentage of students entitled to education maintenance allowance (registered for education maintenance in Scotland)

This information will show for each institution attended by the applicant along with the contextual data for that school/college (where the applicant has selected the school/college from the UCAS list and data is available).

UCAS will also supply a field for the POLAR2 score for the home address provided by the applicant. POLAR2 provides an indication of the level of participation in Higher Education of an area at constituency level.

This data will be imported into CamSIS, but not visible on a screen. It will be at the discretion of each College as to how this data is used.

Moderation spreadsheets

Since the 2011/12 admissions round, the moderation spreadsheets produced by CAO have been enhanced to include six contextual data flags and to provide a preliminary ranking of applications based on the findings of the Multiple Regression Study undertaken by staff in the CAO Admissions & Data Services team under the direction of the Admissions Forum's Working Party on Admissions Research.

Flagging

There will be six flag columns, immediately adjacent to the applicant's name. All applicants will be potentially flaggable. Colleges will be able to update entries. The flags fall into four types:

Flag type 1: in care

• Flag if an applicant has been in care for any duration (at any time in their lives), derived from their UCAS application

Flag type 2: socio-economic

• Flag if an applicant's home postcode is in the bottom POLAR2 category, i.e. they are from a Low Participation Neighbourhood
• Flag if an applicant's home postcode is in one of Output Area Classification (OAC) super-groups 1-5 or sub-groups 7a1, 6c2 and 4c1, i.e. they are likely to be members of groups from which we currently receive fewer than half the undergraduate applications that we might expect.

Flag type 3: school

• Flag if an applicant's current school/college has produced fewer than five successful Oxbridge applicants over the last five years
• Flag if an applicant's GCSE school is in the maintained sector and has a capped GCSE score of less than 40

Flag type 4: CSAS

• Flag if an applicant has indicated on the SAQ that they are submitting an Extenuating Circumstances Form (ECF)

Ranking

Four AS/A2 UMS averages will be generated and presented in four adjacent columns, for applicants who have provided UMS scores for at least three subjects, with a total potential score of at least 600 across the three subjects.

The average of the best three subjects and the average of the best four subjects (where available) will be provided for all such applicants. For applicants to science and technology courses, additional columns will indicate the average score across all science subjects (where for this purpose the science subjects are Biology, Chemistry, Maths [including Further Maths] and Physics) and the average over Maths (including Further Maths) and Physics only. In the first three cases the average will be calculated as the average UMS percentage with each subject weighted equally; in the case of the Maths and Physics average this will be weighted in proportion to the total potential UMS scores available in the units taken for each subject.

Applicants will be ranked according to a Merit Score. The established formula will be used to rank applicants for Medicine and Veterinary Medicine.

Applicants who can be ranked will be banded in quintiles, labelled A-E. Within quintiles, applicants will be ranked by their Merit Score. Applicants who cannot be ranked will be grouped together at the bottom of spreadsheets, with a clear means of identification (they will be allocated to 'quintile' Z).

Merit Scores

In all subjects except Medicine and Veterinary Medicine, each applicant will be placed into one of five groups, as indicated in the 'Rank Group' column:

• UMS: Applicants with UMS for at least three subjects, with a total potential score of at least 600 across the three subjects
• UMS+GCSE: Applicants with UMS for at least three subjects, with a total potential score of at least 600 across the three subjects and at least 7 GCSEs
• SUMS: Applicants with UMS for at least three science subjects, with a total potential score of at least 600 across the three subjects
• SUMS+GCSE: Applicants with UMS for at least three science subjects, with a total potential score of at least 600 across the three subjects and at least 7 GCSEs
• [blank]: All other applicants – i.e. those without sufficient UMS data to be placed in any of the above groups

The ranking formula for each subject will be based on the findings of the Multiple Regression Study, and will be primarily based on UMS data. The Rank Group indicates which model will be used to calculate a Merit Score. In some subjects, UMS and GCSE together provided a better fit to Tripos results than UMS alone: in these subjects, Merit Scores will be based on UMS and GCSE data for the UMS+GCSE Rank Group, and then on UMS data only for applicants in the UMS Rank Group. In subjects where GCSE performance was not found to add value to the regression model, all applicants with UMS data will be placed in the UMS Rank Group, and assigned a Merit Score on that basis, regardless of whether they have GCSE data.

Science-subject UMS (SUMS and SUMS+GCSE Rank Groups) will be used for science and technology Triposes only in cases where the science-subject UMS measure has shown a greater correlation with Tripos performance than the best-three-subject UMS measure. Again, GCSE data will only be used if it improves the model.

Applicants without sufficient UMS data will not be assigned a Merit Score and will be assigned to the Z group in the Quintile column. This is in contrast to the 2012 application cycle, in which all applicants with GCSEs were assigned a Merit Score even if they had no UMS data.

Interpreting Merit Scores

Merit Scores will be displayed over three columns: Normalised Merit Score, Merit Range and Quintile.

• Normalised Merit Score: the Normalised Merit Score shows the Merit Score achieved by the applicant, as a score (roughly) between –20 and 20. A Normalised Merit Score of 0 would indicate an applicant right in the middle of the ranked cohort, in terms of prior attainment, with about two-thirds of applicants falling between –10 and 10.
• Merit Range: this shows a prediction of the likely Part IA/Part I Tripos performance of the applicant, as a range of overall percentage. Based on the performance of previously admitted students, we would expect a student with these UMS/GCSE scores to attain a Part IA/Part I percentage within this range approximately two-thirds of the time.
• Quintile: this divides Merit Scores into five equally sized groups, with A designating the top 20% of ranked applicants and E the bottom 20%. Unranked applicants will be placed in a Z group. Please note that applicants at the bottom of one quintile will have very similar scores to those at the top of the next quintile down.

It is worth noting that there is a significant skew in UMS performance, with many applicants achieving very high average UMS percentages. Normalised Merit Scores will have a natural cap somewhere around a top score of 20 (since UMS percentage can never be higher than 100%), while very low scores are possible where an applicant's UMS percentage is substantially lower than the average for the cohort.

1.4.5 De-selection

The Undergraduate Admissions website states the following in its Interviews section:

‘Almost all of our applicants are predicted top grades so it’s difficult to select fairly based on the UCAS application alone.

‘Admissions decisions at the University of Cambridge are based solely on academic criteria – your ability and your potential – and along with all the other information you provide, interviews help Admissions Tutors to assess your application. 

‘We try to interview everyone who has a realistic chance of being offered a place, which is over 80 per cent of applicants each year. If you have a good examination record and a favourable reference, it’s likely you’ll be invited to interview. However, due to the level of competition for places, there are applicants each year who are not interviewed.’

This endeavours to make clear to prospective applicants that they are not guaranteed an invitation to interview, but also commits the Colleges collectively to inviting to interview all applicants with a realistic chance of being offered a place. Within our decentralised admissions system it is therefore important that Colleges are seen to act fairly and consistently in deciding not to invite individual applicants to interview.

The following guidelines for the making of de-selection decisions have been agreed by the Admissions Forum:

• De-selection decisions will only be made after the following two phases have been carried out:
  (i) careful consideration of prior academic performance and predicted grades alongside the school reference and personal statement
  (ii) the Admissions Tutor has been through the application file to ensure that all relevant contextual information has been taken into account.
• If an applicant achieves less than 80% in relevant AS and/or A2 units and also has a weak GCSE portfolio, then an interview will be unlikely. Colleges may, if appropriate, choose not to interview some applicants with more than 80% in relevant AS and/or A2 Level units.
• De-selection decisions should be made in consultation with the relevant Director(s) of Studies, with the ultimate decision resting with the Admissions Tutor.
• Although their individual circumstances may quite naturally merit special consideration, Extenuating Circumstances Form (Section 2.1) applicants are not guaranteed an invitation to interview.
• Where there is doubt, the benefit should be given to the applicant.
• Colleges should closely monitor their de-selection practices.
• The proportion of de-selected applicants would not usually be expected to exceed 20% of the gathered field (including overseas applicants) across the University, although the de-selection rates in individual subjects may be higher than this.
• Subject moderation should be conducted to ensure that Colleges are making fair and consistent de-selection decisions.
• Admissions Tutors should consult the Director of Admissions if they feel that the proportion of applicants they wish not to call for interviews is higher than the levels that have been mentioned in previous discussions in any subject and/or if this is out of line with their College’s previous practice.

Colleges are encouraged to communicate decisions on their de-selected applicants to CAO for onward communication to UCAS before Christmas. Colleges should endeavour to do this in all subjects except Medicine and Veterinary Medicine by 15 November. Colleges should then write to any de-selected preference/allocated applicants explaining that they will not be interviewed. In some circumstances it may be more appropriate for the Director of Admissions to write to de-selected Open applicants; Colleges should discuss this option with him on a case-by-case basis.  De-selection decisions on applications for Medicine and Veterinary Medicine should not normally be finalised until the receipt of BMAT results. Communication with de-selected applicants (including Open applicants) in these two subjects should therefore be dealt with by the relevant College at the appropriate time.

1.4.6 Offers

In order to assist Colleges in phrasing clear and accurate conditional offers to applicants, a guidance document is given in Appendix D. Further information can be sought, if necessary, from Helen Reed in CAO.