Why a data-driven culture in education matters

The roll-out of innovative data-driven solutions in the Belgian education system is not something that happens overnight. The heavy workload of most education staff, caused by a major personnel shortage, turns data analysis and usage right into a challenge. And even more so, into a concern. Our Data Scientist Hendrik De Winter therefore exposes the three most common doubts with some tangible answers.

Curious? Let’s dive right in!  

Are you ready for tomorrow?



The digitisation trends 

Every academic year, teachers, professors and school systems strive to improve the way knowledge, skills and attitudes are transferred to students. The introduction of new technology at various levels of learning, from our youngest students to bachelors and masters, has led to greater efficiency, more consistent decisions and a shift towards individualised learning. 

The administration, communication, and assessment aspects, which were always part of the teachers' job, have gotten faster and smarter. For example, they now take the form of custom-made web applications and portals such as Smartschool, Toledo, I-learn and others. On top of that, analytics and other innovative, digital solutions are emerging across the Belgian school system, such as audio-visual databases for teaching materials, road safety lessons in virtual reality and a digital skills passport to help with study choices later on in the curriculum.

Quite awesome, don't you think? 

All the new solutions resulting from digitisation have directly led to better monitoring of students over time, across gender, ethnicity, country and fields of study.

All the new solutions resulting from digitisation have directly led to better monitoring of students over time, across gender, ethnicity, country and fields of study. Moreover, they contributed to a better understanding of ever-evolving realities.

Databases of education statistics and indicators emerged, propelling the sector forward with useful and accurate insights about the school population.

Furthermore, administrative processes within schools (think of the long queue of parents to enrol their children in some schools in Belgium and the recurring purchases of textbooks every year) were also improved as a result of more intelligent tools.

Additionally, they also offered the possibility of providing feedback on teachers and teaching strategy and vice versa. In short, it can be said that the learning experience in general has changed dramatically. However, there is still a large number of applications of new technology and specifically data-related to explore as it is today. Also, not all ideas stick immediately and dissemination of these solutions takes time. 


The challenge of innovation adaptation 

The introduction of innovative data-driven solutions in the Belgian education system, as in other areas such as the healthcare and public transport, does not happen overnight.

The heavy workload of most school headmasters, staff and teachers, caused by a major staff shortage, makes the analysis and best use of this heap of new information (unfortunately) a challenge. The belief in the power of innovative, digital solutions versus traditional methods is also not uncontroversial within today's schools.

Moreover, there is often a knowledge gap that prevents schools from experiencing the true potential of a shift to data-driven and informed working. Creating awareness about what impact innovation and data can have on growth opportunities within the school walls is still extremely imperative.

Although, besides creating awareness, it is equally important to address the recurring concerns about data-driven education through some clear value-adding examples.

In what follows, we therefore highlight the potential role of data within education based on some typical concerns that often overshadow this topic as well as our responses to them. 


Typical concerns and why they don’t apply 


Concern No. 1: Data is dehumanising 

"Nobody wants to be reduced to a number or a set of statistics." 

The positive impact of data-based education is broader than just the "standardisation" of tests and the ill-considered use of grades to describe the students’ progress. Today, the Belgian education system is based on what we call a bell curve. Every year, the curriculum is drawn up for an average group of people, representative of every pupil in Belgium. Consequently, feeling ahead in some subjects and behind in others is quite normal.  

MicrosoftTeams-image (156)



Data and its applications, think of a predictive (data science or AI) model, can help to achieve an adapted and individual approach.

As a result of segmentation, fast advanced learners are identified and served properly, and more creative learners are challenged differently. Personalised learning relates to the level and difficulty of learning, but also to the way it is delivered using images and videos, with or without many intermediate tests, through classroom teaching or in group discussion.

This change is expected to lead to a big increase in the motivation of a class of students, better retention of learning material, and thus ultimately a big increase in performance. 


Concern No. 2: Data has never been proven effective 

“We cannot associate data-driven decisions with student learning and achievement” 

One of the goals of our school system, especially in later stages, is simply to help people graduate. Moreover, another goal is to constantly look for ways to increase graduation rates and reduce dropout rates.

It is true that sufficient studies finding support for data-based learning (rather than more traditional learning methods) have not yet been performed.

However, we believe this is only a matter of time, as there is so much in it for both student and teacher, if used intelligently.

For instance, after the academic year, student data can help teachers gain valuable expertise by precisely analysing, monitoring and understanding which teaching methods worked well.

In addition, it can help identify where there is room for improvement, and whether the timing of the course may have been scheduled too early in the morning or too late in the afternoon, potentially losing concentration.

School infrastructure, availability of teaching materials and use of class time are known to affect learning outcomes. This assessment will lead to a better learning environment year after year. On the other hand, students themselves can benefit from tailor-made courses in smaller groups, test results communicated to them individually and the overview of their own performance dashboard and acting based on it. 

Data helps identifying students likely to drop out, and even predicting them, in order to proactively support them with additional resources and materials.


For teachers, manually striving to follow up students individually is obviously possible in a smaller school. In a larger school with a large group of pupils, this is much more difficult.

At the level of secondary schools and Belgian universities, we therefore once again rely on the help of data and algorithms, which already teach us today which factors influence performance, lack of performance and loss of motivation.

Data helps identifying students likely to drop out, and even predicting them, in order to proactively support them with additional resources and materials. 


Concern No. 3: Data makes our teachers' lives even more difficult than they already are. 

"The extra weight of data-driven learning is not justified" 

Teachers and school headmasters concerned about the extra workload and cost of data are not alone. However, it is important to introduce all stakeholders to the benefits of intelligent data-driven working. To demonstrate this, we review some examples of automated assessment and other educational support: 

  • Example 1 
    Automatic assessment of multiple-choice questions has been around for some time, but now AI also helps correct answers to more complex questions (like open-ended ones). Think of correcting spelling in language courses, identifying key words and checking plagiarism. 
  • Example 2
    When grading online assignments, intelligent platforms already learn what mistakes are often made within a class. They consequently alert the teacher on this. The result could be a reminder to formulate a question better (because it is misunderstood) or an indication that the question might be too difficult. 
  • Example 3 
    When assigning homework online, analytics can be used to evaluate, for example, the opening rate of the task, scrolling behaviour on the page and amount of time it took. This leads to an easier tracking of the students.

  • Example 4 
    Using various tools, student attendance and patterns in absences can be better recorded and monitored. In general, repetitive tasks can also be automated. 


Fancy knowing more about your data-driven (education) approach?  

Hendrik De Winter is our Data Scientist, active in the field and working at Tobania since one and a half years. As recent graduate, his interest in using data in educational environments stems from personal curiosity and testimonials from young, bright teachers around him.

Want to know more about his article or about our data offerings like for example Data Foundations, Data Strategy, Data Science, and Business Intelligence in general? Then be sure to contact Hendrik and our full experienced data team. They are always happy to help:  


Hendrik De Winter

Hendrik De Winter