As a child, I used to love playing animal, vegetable, mineral - the game in which someone thinks of an object, tells you the category, and then you ask up to 20 yes/no questions to identify it.

Each question revealed a little bit more and responses would require you to change your next question.

The volume of personal data we collect on our students at enrolment is mind-boggling, yet of all the sectors in which I have worked - local government, retail, insurance, financial credit, electricity generation and education - I think educators know the least about their customers.

Understanding students

We also tend to ask lots of questions at enrolment and very few - the occasional survey aside - afterwards. So it is not surprising our knowledge is so poor.

Our practices would be transformed if we ignored the data that the government says we should collect (while also dropping the pretence that the government only asks us for information we’d collect anyway to run a successful college), and instead focused on getting information that would help us understand our students and provide a more tailored service.

So what might we ask to properly understand our younger students? Here are some suggestions. If I could only ask two things to help me predict whether a student would be successful, they would be:

• “Are your parents pleased that you are coming to college?”
• “Did your parents always attend parents’ evenings at school?”

Parental support correlates very highly with student success. If mum and dad take an interest, you are more likely to succeed; if they don’t, that alerts us to a need to monitor you more closely and support you early on.

Prior attainment matters

Another key question is, what grade did you achieve in GCSE maths? Last year, I asked my health-economist son to do some multivariate analysis on achievement. Within a few minutes he pointed out an almost perfect correlation between maths GCSE grade and achievement in other subjects. Anyone with a B or better (in old money) always passed; at the lower end, that fell to 50 per cent.

We had never spotted this because so many at the lower end left within the first six weeks (such students are regarded as never existing). The answer gives you a clear pointer to where you need to devote initial attention.

Then questions such as “Do you have a part-time job?”; “Did your parents go to university?”; and “Do your parents vote in elections?” enable you to tailor things such as work experience, guidance and British-values elements. The guide we send to parents explicitly states that we expect them to support us on the values agenda by voting at elections.

Finally, once students are settled in, there are more pieces of important information we can glean. These include areas such as home access to the internet; preferred method of contact (email is for old people); asking about the scariest thing you have ever done; how many other people at college you already know (loneliness damages outcomes); and what time you go to bed (lack of sleep can be an indicator of serious health problems).

Who needs most support?

These questions quickly give you a picture of where to focus effort as an educator, and maybe where you have been wasting effort. They tell you which students might need support to make up for a lack of it elsewhere.  They help avoid reasonable criticism, such as “Stop preaching about what employers want, I work 10 hours a week for one already.”

Asking questions needs to become second nature, too. We introduced contactless payment in our canteens without ever thinking to check whether our 16- to 18-year-olds had debit cards or could make contactless payments, for example.

The use of big data in both good and immoral ways is constantly in the news, be it curing cancer, Cambridge Analytica or the Brexit campaign. Google can predict epidemics before health services. Facebook constructs an accurate digital version of you based on everything you do on the site.

But further education is still in the dark ages. Most of our data experts seem to spend their lives looking at PDSAT reports and correcting errors rather than acting as data scientists providing stunning insights about our student population so we can help them and improve our performance.

Better data

We need to shift the dial, collect much more useful data than we do, and use it. When you see colleges asking questions to gauge student satisfaction before they’ve done the research on what types of things are important to a student’s satisfaction, you realise how amateurish our approach is.

Our students are much more than their gender, ethnicity, postcode, National Insurance number and exam results. We are doing them a huge disservice if we simply collect what we’re told.

Every one of our students is unique, with their own reasons for coming to us. They want to achieve something important to them. They are a precious commodity. Let’s take a proper interest in them as individuals and make sure they know how special they are, and create an individualised learner record worthy of the name.

Ian Pryce is chief executive of Bedford College