Selective admission

It sounds like a common sense policy: programmes with a limited number of spots prefer to give those spots to the best students. It’s also a financially sound policy: the best students study faster and have lower drop-out rates, which is good for the programme’s reputation.

An increasing number of programmes applied for a numerus clausus policy in the last few years, and with good reason. This policy allows them to practise selective admission. But how does one select the best students? Or rather, the most successful students?

Susan Niessen has spent years trying to answer this question. Her research focuses mainly on the psychology programme, which has been experimenting with selective admission for years. But she says her results can inform many other programmes.

Mini exam

One useful method for predicting student performance is no longer allowed. High school academic performance is a pretty good indicator for university academic performance, the  but the law now forbids programmes from selecting students on this criterion alone. So what criteria can they use?

The method of test studying worked much better to predict a student’s performance than others

The solution turns out to be simple. ‘A few exams are usually enough to show whether someone will be a good student’, Niessen says. ‘Study programmes should administer these as early as possible.’

In other words: students should sit an exam before they ever start classes.

The programme implemented the practice of ‘test studying’ in 2013. ‘Students who registered for our course are given first-year material to study’, Niessen says. ‘After having studied it, they have to come here to sit an exam.’

In the years since, the programme has added a mini exam in statistics and a lecture. Niessen recorded students results and then followed the progress of each batch of students.

Mums and dads

The results were conclusive. The method of test studying worked much better to predict a student’s performance than others – such as writing a motivation letter, for example. In the past, the psychology programme has required students to write one.

‘It’s completely useless’, says Niessen. ‘It doesn’t work as a predictor at all. In part because mums, dads, and brothers and sisters usually help out with writing the letter. But the programmes have no way of knowing that.’

Study programmes also attempt to measure non-cognitive characteristics: whether students know how to keep a schedule, or whether they’re sociable people, or honest. ‘In the United States, they even check for leadership qualities, or active citizenship’, says Niessen.

‘But they try to measure that through questionnaires’, she says. ‘People have to check whether or not they think of themselves as “hard workers”, or a “good planner”.’

Socially desirable

It’s easy to see how these questionnaires can be manipulated by people only checking what they perceive to be socially desirable answers. During a selection process students may want to present themselves as better than they really are. ‘It turns out students were indeed doing this’, says Niessen. ‘The predictive value of questionnaires like these in a selection process is basically zero.’

Interviewing prospective students is not only too time-consuming to do on a large scale, it’s a poor indicator of study performance. ‘An interview also allows students to only give socially desirable answers’, she says. ‘And the final verdict is always subjective.’

If you do have to make a selection, be as thorough as possible about it

Nevertheless, the psychology programme’s selection methods are on the right track, says Niessen. The key is to select on the basis of what the programme actually involves. Does it involve a lot of literary theory? Do students have to work independently? Programmes should test prospective students on these aspects during selection. Do other aspects, such as social skills, influence study performance? Programmes should test for these during selection just as they do for their curriculum.

Marginal

But test studying is not a perfect solution, says Niessen. ‘It’s not like people never make the wrong decisions’, she says. Critics of the method say there is very little difference between a number 500 on the list and number 501. This would mean that it’s unfair to admit number 500 but reject number 501. But, she emphasises, ‘there is a difference between number 300 and number 500. So if you do have to make a selection, be as thorough as possible about it.’

And clearly, selection only works when programmes actually have the opportunity to choose. ‘In reality, students applying for a clausus programme are rarely rejected’, says Niessen. ‘This means the effect on the study output is marginal. But if a programme has more application than spots available, it can be a useful tool.’

Dutch