Using UCAS data to identify language trends 1

Late last year I wrote about the difficulties involved in calculating how many students are studying languages, by using HESA data. An alternative source of data is from the UK’s University and Colleges Admissions Service (UCAS). One of the great advantages of the UK centralised admissions system it does mean access to some good data. It is possible to find out which courses are in the highest demand (by looking at application to accept ratios), details about the socio-economic profile of students, school grades etc. Arguably, UCAS data is presented is much more user friendly way than HESA data.

Caution about UCAS data

The case may be different for other subjects, but UCAS data is not very reliable for tracking trends in individual languages. The subject data presented by UCAS is based upon an applicant’s primary choice—in short UCAS staff look at an application form to see which subject the applicant is primarily applying which then forms the basis of the subject choice

Despite the clarity of the layout, there are at least two reasons why UCAS data is not particularly reliable for mapping trends in individual languages.

  1. There is very little allowance for joint and combined honours combinations to be recorded. Where they are reported, it will be where the applicant has expressed no clear preference for a subject—these students are recorded in Y Combinations of Languages with arts and humanities, social studies, combinations of languages. We do not know what combinations they are actually doing. Unspecified combinations of languages attracted 9,589 application and 1,652 accepts. Add to this the combinations of languages with arts/ humanities (another 7,623), languages with social studies, business and law another 2,589. This is considerably more than the 4,678 studying European languages (Group R) and the non-European languages and languages (1,485).
  2. For individual languages ‘trends’ are extremely volatile, probably not because demand is volatile, but because classification practices vary year on year. This is especially the case of ‘less widely taught languages’. Taken at face value it seems that only two students applied to study Portuguese last year.

The upside

However, as each individual is only counted once we can work out how many people might be studying for a language degree in higher education whihout worry are tenths, third and halves of students. If we add up the Group R total, the Group T totals and the Y language combinations for 2010 we get 18,027. Multiple that by 4 (assuming all these students are doing 4 year courses) and we get 72,108. This may be an underestimate as I have not included the sciences with languages/ arts and humanities, but it could also be an overestimated as I this 72,000+ assumes four-year degrees, so if I was to multiple by 3 we get 54,081 not especially close to the 42,444 I estimated from the HESA data.

UCAS is about admissions and applications to study, not statistics per se. It tells us nothing of the numbers of students who enrol on language courses in Institution-Wide Language programmes or change to (or from languages on arrival), though some of the these students are recorded in the HESA data (it depends though on how institutions report their data to HESA.

The data on which this post is based is on the UCAS website. I will post the past five years of data from the individual languages in the next day or two.

  • Twitter
  • Digg
  • Facebook
  • Technorati
  • Reddit
  • Yahoo Buzz
  • StumbleUpon

1 thought on “Using UCAS data to identify language trends 1

Comments are closed.