The CSSPS algorithm allocating schools to students in Ghana
These days, September has become the month where there’s so much confusion around how thousands of Junior High School (JHS) students are not placed into Senior High Schools (SHS) in Ghana.
At the centre of this annual confusion is the CSSPS algorithm which the Ghana News Agency (GNA) captured in the quote below:
Some parents in an interview with the Ghana News Agency (GNA) expressed their frustrations over the challenges they were going through over the inability of the Computerised School Selection and Placement System (CSSPS) to allocate schools to their wards.
Source: GNA
There’s a lot to unpack here: the CSSPS algorithm, parents, students, and a lack of information.
Recently, Kehinde Ajayi, Willa Friedman and Adrienne Lucas did a really interesting study around these issues. It an intersection of students, parents, information and the CSSPS algorithm.
What they did was to provide parents and students with detailed information such as the WASSCE pass rates of SHS Schools, the average score of BECE candidates accepted and much more. The idea is to see how such detailed info affected the school students and parents selected and what it meant for the kind of schools the CSSPS algorithm placed students.
In the end, they found that information changed the characteristics of schools which students selected as their choices.
Beyond their final results, what I found really interesting was the details they shared about the CSSPS algorithm that allocates schools to students. I’ll share my thoughts on some of these details from the study that surprised me a bit.
52% of students admitted to a school on their list of school choice
The idea is to give your school preferences, grades etc to the algorithm, with the possibility of it being able to match students to their preferred schools and it does that just half the time?
This surprised me a bit because it could mean many school choices students make are similar and so the algorithm is forced to match them with schools outside their choice. These days, I learn the number of schools students can choose has been increased from 4 to 5, so, that may solve this problem slightly.
The algorithm thinks in raw scores
I found this interesting because I never really thought about the possibility that the algorithm uses raw scores and not the final aggregate.
For instance, while 6 1’s may seem the same, they are not, because, to the algorithm their raw scores are different. We’re used to discussing results in grades probably because that's how final results are reported.
By implication, high raw scores are ranked higher than lower raw scores, regardless of the economic background of the student. However, the algorithm partly solves for this problem by providing a “catchment area” option.
Private schools are ranked between Category A & B schools
They found Private schools to be ranked between Category A & B schools. This ranking was done using a metric they created called value-added. It compares BECE scores of accepted students to WASSCE pass rates. The wider the gap, the more a school’s value-added. This provides a better sense of where to place private schools as sometimes they can be assumed to be better than public schools simply because their fees are higher and by extension, provide a better chance of doing well at the SHS level. I wonder where Technical/ Vocational Schools rank in this.
Conclusion
The study noted that some students clarified how the information received came after they or their parents had already decided on their school choices which means it was probably too late to play a major role in their decisions.
However, I still think the bridging this information gap is an opportunity for Junior High Schools to engage parents and students on an ongoing basis to help them make more informed choices about alternatively-good schools outside the very well known and competitive ones.
The study has so much more interesting details around how information affects student aspirations, how selection strategy information affects choices and how the algorithm works.
You can find the work of Kehinde Ajayi, Willa Friedman and Adrienne Lucas - When Information is Not Enough: Evidence from a Centralized School Choice System - here.
Thanks for reading!
