The weathered-steel facade of the Amsterdam University College building

Fair maths admissions and higher pass rates in first-year statistics

45
countries represented in the student body
2018
using Sowiso since

“Now we can compare students from different educational systems that we don't know much about.”

Kasia Malarek, Admissions Officer and Lecturer, Amsterdam University College

Amsterdam University College (AUC) is a selective liberal arts and sciences college. Its students graduate with a joint Bachelor’s (Honours) degree from the University of Amsterdam (UvA) and Vrije Universiteit (VU) Amsterdam. The programme follows the American liberal arts and sciences model, all teaching is in English, and AUC admits a maximum of 300 students a year. Its student body comes from 45 countries.

Maths matters at two points on a student’s path through AUC, and Paragin Group is involved at both. Before students are admitted, OMPT gives the admissions team a clear view of the maths level of applicants whose schooling is hard to judge on paper. Once they are in, Sowiso supports the first-year statistics course taught by Dr Michael McAssey.

The challenge: maths levels from 45 countries, and a statistics course without a fitting textbook

Admissions: comparing maths across school systems

AUC offers three majors: Sciences, Social Sciences and Humanities. Each has its own maths requirements, and the admissions team has to decide whether an applicant meets them. With applicants from so many countries, that means comparing more than diplomas. The team looks at each applicant’s individual maths programme, while curricula, grading cultures, assessment methods and grade distributions differ widely from one country to the next.

Kasia Malarek, Admissions Officer and Lecturer at AUC, describes two groups of applicants. The first comes from large, well-known systems such as the Dutch VWO, the International Baccalaureate, the European Baccalaureate and A-levels. The second comes from less familiar systems whose maths programmes the team cannot assess. For that second group, establishing an applicant’s maths level was complicated.

“We tried to understand the educational system we were working with. This takes a lot of work and changes on a yearly basis. So it was tough to compare.”

AUC wanted an objective way to keep its selection fair and dependable.

First-year statistics: a course that never quite worked

Dr Michael McAssey has taught at AUC since 2013: Statistics for Sciences, Basic Research Methods and Statistics, Advanced Research Methods and Statistics, and Calculus. As a member of the Academic Core Faculty, he is in charge of Basic Research Methods and Statistics, a core course with around 60 students per semester, taught in classes of 20.

“We had a long process of the class evolving over time, and we couldn’t reach a point where we were satisfied with it. Number one, we couldn’t find a textbook that we liked, that was at the right level, and that had examples and exercises in the discipline the course was directed towards.”

Michael and his colleagues started with ungraded homework, asking students to check their answers against the solutions in the textbook.

“Obviously, a lot of students didn’t do that and didn’t do well on their exams, so then we wanted to grade their homework, but there were too many papers to grade, so that didn’t seem feasible. And, of course, there was also the issue of students not wanting to spend more than 100 euros on a textbook that they were only going to use for one semester before reselling. So we wanted to get away from having a textbook as well.”

Next, Michael tried an online platform that graded assignments automatically. It saved time on marking. But every student got exactly the same questions, so students copied from each other and did not really learn the material.

The solution: proctored maths exams from home, and randomised exercises with feedback

Admissions: OMPT where a maths level is hard to judge

AUC has used OMPT in its admissions for several years, for applicants whose maths proficiency is difficult to evaluate. It works with three exams, each aligned with the maths a major demands: OMPT-A for Social Sciences, OMPT-B for Sciences and OMPT-C for Humanities. Applicants anywhere in the world can take the exam online.

“What is great about these exams is that the applicant can take a test in the comfort of their own home. But also because it’s proctored, it assures our results are accurate.”

OMPT also offers preparation materials and mock exams, so applicants can improve their maths and arrive well prepared for the exam and for their courses. AUC encourages applicants to use them.

Candidates usually pay for OMPT themselves when they apply to other institutions. One year, AUC decided to cover its applicants’ exam costs, before it had assessed whether they were eligible, and set an earlier deadline of 1 July for the maths exam. That took the financial burden off applicants and gave the admissions team a better view of eligible candidates.

First-year statistics: the same question, different numbers

For his statistics course, Michael knew what he was looking for. He found it through a student.

“So what we really wanted was a system that would ask students essentially the same questions but with different numbers. This was when we came across Sowiso since one of our students was doing an internship there.”

AUC started with Sowiso in 2018. Michael could now set assessments in which every student got the same question with different variables. Because all the content is online, students no longer had to rely on a printed textbook. The Sowiso team also adapted the content to what AUC wanted to teach.

“The staff at Sowiso are willing to listen to us and make changes and add things to the content so that it’s covering the things that we want to cover and in the way we want to cover it.”

Automatic grading and randomisation were what drew Michael to Sowiso. Over the years he has come to value the automated feedback just as much. Sowiso analyses each answer, identifies where the mistake is and what it is, and gives hints that fit the student’s route to the solution, after a wrong answer and after a correct step alike.

“Unlike doing exercises in a textbook where you might not even know if you were right until you review it with someone else or you look up the answer in the back. And even then, it’s only the final answer, so you don’t see how to get to it.”

The result: fewer students failing statistics, and a fair way to compare applicants across school systems

In the statistics course, the change showed first in how students worked. Michael saw that the automated feedback and the interactivity of the platform made students more motivated to engage with the material, and passing rates went up.

“The students enjoy the online platform and appreciate that they can use it instead of purchasing an overpriced textbook. They are also having more fun with it. They can work through examples and practise and get feedback on their answers, and we also have set up the assessments so that they have a chance to correct errors before submitting their work.”

For Michael, the platform gives far more flexibility than a static textbook to shape the course the way he wants. And, as he puts it, “we have less administrative work to perform so that we can focus more on our teaching activities.” In 2020 he planned to bring the same approach to his Statistics for Sciences course.

When Sowiso followed up with Michael in 2023, the platform had become part of how the course runs. He uses it for assignments and tests, and for practice outside class, and assignments and practice exercises count towards the final grade. Teachers have noticed that fewer students fail. Michael believes Sowiso motivates students partly because it gives them autonomy: they use the material in a way that suits how they learn, and they learn from their mistakes.

“Before using Sowiso, we had to grade everything ourselves, which is very time-consuming. So we picked a sample of exercises to grade, which still took a lot of time. Students don’t want to do exercises that will not be graded. So they were not practising enough in the end. But now, with Sowiso, we include everything in the student’s grades. For Statistics for Social Sciences, the grades went up since students were doing all the work and more practice. This tool improved their performance in this course since, thanks to the randomisation, they could do more exercises than before.”

Students notice the difference themselves.

“They like it very much. They can go at their own pace. This is something that they mention a lot.”

Michael’s next goal is to keep improving the course design with Sowiso.

On the admissions side, OMPT gives AUC a shared reference point for applicants whose schooling it cannot assess on paper. The pass mark is 75%, and applicants get a maximum of two attempts, so that one bad exam does not decide their application.

“We think this gives applicants a nice opportunity not to be very stressed about their exam and to have in mind that they have their second attempt.”

The admissions team saves time assessing qualifications, and its decisions are fairer and more consistent. The detailed results also tell AUC something about how an applicant is likely to do once admitted.

“Now we can compare students from different educational systems that we don’t know much about. We can predict how well they will do in the programme because, together with the results of the OMPT, you get more detailed information, showing how the candidate did in more particular aspects of the exam. So we have a great indication of the general and specific skills required to complete each major successfully.”

← Back to all stories

See what this could look like at your organisation.

Talk to us about your assessment, from secure exams to certification.

Don't miss these

Making mathematics more accessible to non-math majors
Building continuous assessment into first-year mathematics
Online maths entrance exams and accounting pass rates of 75-80%
A decade of innovation, and what an AI solution bot changed