Photo by Levi Saunders on Unsplash

Yesterday was July 6th, the day I finally got my IBDP (International Baccalaureate Diploma Programme) results. It was when I would finally find the outcome of my efforts and hard work that I put in the last 2 years. I was nervous for the results and excited for a fresh beginning that would begin in the university. All of my friends were curious to know their results, and their parents were hoping that their child has not wasted an immense amount of money and time.

At 6:30 pm IST results finally arrived. My predicted grades were 40/45, which is considered pretty good for IB, and I got 38/45, which is still pretty good, and I have no problem with it.

The Problem

So first I’ll explain to you my problem and then a general problem.

The issue comes when we look deeper at the results. I got 7/7 in Math and Physics, but honestly, I don't deserve it. I would have gotten 7 in Math if we had exams. In physics, I would have gotten maximum 6, but mostly 5. In Computer Science, for which I worked hardest and only once got below 7 in 2 years of school. On that subject, I got 5.

Now when university will get these results, they would give me credits for subjects I don’t deserve and I won’t be given credit for the subject I know. This would mean that I will have to study again few things that I already know and I need not study the things I don’t know in Physics and Math.

This is just fcing insane. That means I will have to pay hundreds of dollars for unnecessary classes.

There are many who are in a way worse situation than mine. Some even got 12 points lesser than their predicted grades. Their universities might not consider them at all for admissions, they might have to enroll again for a new session, and many might quit their further education.

Why should students pay for some stupid error in the statistical model created by IB?

How This Catastrophe Happened, and What IB Is Doing to Evaluate Students

The pandemic has led to a lot of changes in operations of IBDP. Because of the pandemic, IB had to cancel its May 2020 examinations, and they had to find an unconventional approach to grade these students. They decided to use statistical modelling to come up with ‘accurate grades’.

meme from imgflip

The video above is by IB and explains their new approach of assessment, and below is the text version from their website.

The IB will be using historical assessment data to ensure that we follow a rigorous process of due diligence in what is a truly unprecedented situation. We will be undertaking significant data analysis from previous exam sessions, individual school data and subject data. Our IB staff are working with an education organisation that specializes in data analysis, standards, assessment and certification. Together we have developed a method that uses data, both historical and from the present session, to arrive at the subject grades for each student. In our usual way, these grades will be converted into points that contribute to the student’s overall Diploma Programme results.

Basically, these are the metrics for evaluation-

1. Coursework Grades: Grades for projects and assignments submitted by the students.

2. Predicted Grades: The grades that teachers believe a child would have gotten if the examinations were conducted.

3. Miscellaneous data: this includes data such as previous exam data, individual school data, subject data, etc.

   All models are wrong, but some are useful George Box

Why Evaluation Metrics Are Not Fair

1. If a student performed badly in her coursework, she will be penalized twice; Once, because the final grades take coursework grades into account and the second time because of the predicted grades which is also used to calculate final grade.

2. Pygmalion Effect is a real thing. Teachers tend to express lower predictions for the student from the underprivileged background, also teachers may have their own biases for students.

3. Schools that cater to underprivileged communities usually have fewer tests, which means they will have less data to evaluate when compared to richer students. Fewer tests also means that their predicted grades will be less accurate than the students in a school with higher tests.

4. IB said that they would also use the school’s historical data for evaluation. This means that school with a record of high achievers will give better results to the student.

5. IB said that they will include additional data ‘where it is available’. This may include biases because for some students and school certain data may not be available.

There are other issues with the IB grading system that I have not discussed here. I have linked a more detailed article in references.

IB says that the results will be close to accurate results, but we know that statistical models are not reliable and are prone to errors. Even if we consider that IB’s model is 95% accurate, it is still playing with the lives of so many students who spent their invaluable time and money to study in IB school.

It is not fair to let a failed statistical model decide the fate of hundreds of students who have worked their A** off for 2 years

I agree that these are unprecedented times and there is not much IB can do, but at least they should have given predicted grades to the students instead of using historical data and all other kinds of data that may bring noise to the statistical model and play with the lives of students.

What do you feel IB should do? let me know in the comments.

References

160k+ high school students will only graduate if a statistical model allows them to tl;dr : Due to curriculum disruptions the International Baccalaureate (IB) is going to use a statistical model to…positivelysemidefinite.com The assessment and awarding model for the Diploma Programme May 2020 session The IB has provided schools with further detail about how it will assess students' work for the May 2020 Diploma…ibo.org