No Current or Past Mass Intervention Will Revolutionize Education
Forget about the policies and Ivy League Universities
A good friend of mine recalls his shock after joining one of the national schools in the country.
In Kenya, a national school is the top-tier institution known for producing top performers in regional and national examinations. They are given supreme powers to select students who performed exceptionally in the final primary school examination, so the narrative goes.
What he discovered was that most of the top students, those who glossed the top 100 list in the national examination, were not as smart as they were thought to be. As the years passed, his theory has stood the test of time.
Now, not all of them fall in this category. Some persist and continue to dominate in academics. But these are a small cohort, extreme outliers. The obvious impression is that these students cheated their way into the top schools. When you grow up knowing that validation and praise come from having a good grade, cheating becomes incentivised.
In my third year of high school, the government issued a decree to increase the number of national schools. Reasons? I can’t tell. My school was listed as one that was due to get such a promotion. Everyone was ecstatic.
St. Joseph’s School Rapogi, my alma mater, was known for giving national schools a run for their money. As far as academic performance goes, we considered ourselves a national school long before that decree.
However, the local politician didn’t want to hear anything about it. He crossed our school off that list. The reason, so the story goes, was that the local students who benefited from the school’s history of inducing discipline and stellar performance would not get a chance to be enrolled.
He was right. National schools have an obligatory quota from which they have to select students from all over the country. Provincial schools (now known as extra-county schools) would largely select from their locale, their province. And that’s how Rapogi preserved its status.
This is not a mourning session. It clarifies two beliefs. On one end, teachers and parents will do anything to enroll their students in national schools, the alleged top-performing institutions. The other believes that a top-performing school will always perform well, provided it keeps its traditions.
In reality, no past or current mass intervention has ever or will ever revolutionize education.
Interventions
Let’s look at some of these interventions.
1. Incentives — parents will be convinced that a teacher who has received the proper incentive will improve the educational outcome of the students. A fat pocket will only incentivize the teacher to increase the chances of their students scoring good grades. It includes cheating. A strong grade is not a reliable signal for intellectual superiority, as my friend discovered when he got into the national school. Students, too, might not benefit from being incentivized to score better grades. They will only develop smart ways of evading detection when they cheat, which could be used as a form of intelligence when you think about it. But it hasn’t revolutionized education.
2. After-school programmes — these hardly benefit the students. The teachers are the ones mostly motivated because it comes with extra income. A good number don’t care less about the students. If they did, they would have done it at a meagre stipend, or even better, for free. Furthermore, these interventions are used for struggling students. Most teachers already have a fixed mindset about this category. After-school programmes wouldn’t help much.
3. Psychological or behavioural interventions — this may look promising, but we have to remember the evidence used in their favour — psychology. Psychology is experiencing a replication crisis. What this means is few, if any, of their studies can be replicated. Their foundations are shaky at best. It reflects in the interventions used. They are highly variable, and could not be argued in favour of changing educational reforms.
4. Personnel development — this intervention has significant variability. When a teacher advances, they wouldn’t want to remain in the same institution or at the same level. The incentives will push them into something bigger with a fatter cheque or with wider potential. That’s how my class lost its teacher of agriculture. That’s also how my interest in the subject nosedived.
5. Increased resources — in systems dynamics, the rule is that a system is only as strong as its weakest point. If the weakness is not addressed, improving other components hardly converts the system into a robust one. Increasing resources is an assumption that resources are the problem. It isn’t. This, too, will hardly improve education.
6. Feedback and progress monitoring — this shows a lot of promise, especially because it models how systems build resilience and persist in mercurial environments. However, this too is highly variable, likely because it lumps all individual abilities into one and uses a single aggregate tool to determine progress.
7. Tutoring — Here’s where I would put my money. Tutoring has been extensively studied as a way to improve individual learning, with a one-on-one session laurelled as the recipe for creating geniuses. As much as there’s reliable data to argue in favour of this type of intervention, its goals do not align with the powers that be. Educational institutions will lose their business. Ivy League institutions would lose most of their funding, which comes from alumni donations. The status that comes with being an alumnus of a prestigious institution will crumble. All the while, the population-to-teacher ratios are not compatible with this style of study. Multiple vectors are pulling in different directions, which prevent such an intervention from being implemented, most of which are selfishly pursued.
What I have left out, but have already hinted at, is the push to take your kids to top schools. The effects don’t matter much other than the name. Status opens doors. It’s not about education.
My mother used to tell me that she didn’t struggle to take me to a top school so I could learn. She wanted me to interact with the children of the top leaders, entrepreneurs, magnates, and ‘movers’ of society. She understood early enough that education played second fiddle.
The table below simplifies the interventions previously tried to revolutionize education in low socioeconomic societies. But as I have written before, the retention rates are worsening. Thus, even with the best forms of intervention, poor retention nullifies the best of interventions.
The wider the line, the greater the variance, the less reliable the intervention. The reverse applies. Also, the further the red line is from the vertical dotted line, the stronger the reliability of the intervention. Any that is close to it or crosses it has little significance. Thus, of all the studied examples, tutoring ranks as a good option. But who will implement it?
AI will not revolutionize education
I repeat the same mistakes, and so the trouble stays
— Omen
Now that AI is here with us (for how long, though?), the same story has resurfaced. Based on what history tells us, there is hardly going to be an intervention that surpasses one-to-one tutoring. The AI experts know this.
Thus, they have tried to develop ways of tutoring students to imitate the one-on-one training. While this is a commendable step, which might just as well borrow from the outlier effects of feedback and progress monitoring, it overlooks an important aspect — individual effort.
As long as the student is not interested, they will hardly make progress. Yes, most of the household geniuses had tutors, but they were equally interested in the topics. They even challenged the tutors. Anaximander challenged Thales’s hypothesis on where the Earth stood in the solar system. Russell eventually disagreed with Alfred North Whitehead. Kepler improved Tycho Brahe’s ideas on celestial bodies. The changes stem from the details.
However, if AI generalizes what it surfs, it is likely to miss out on some of these details. The other day, I asked ChatGPT to integrate the meaning of Goodhart’s Law, a simple feat. I then asked it whether replacing ‘measure’ with ‘ideal’ would still preserve the interpretation. The black bullet is still pulsating.
It reminds me of the moments we would use calculators to get the square root of a negative number. The result was always a syntax error or a math error. In reality, you can get the square root of a negative number, an imaginary number. This was the story behind the origin of irrational numbers.
What’s more, every time we subject ourselves to AI to confirm our outputs or bursts of creativity, we lose bits of our agency. Thinking does not evolve. Dependency does. That does not revolutionize education.
AI is no different from the previous histories of new, hyped revolutionary educational interventions. Thomas Edison remarked that the motion picture would revolutionize education. It didn’t. Edison was a businessman. The motion picture would likely have expanded his business, just like it would the AI enthusiasts.
Later, radio was announced to be the intervention we have been waiting for to upend our education. Students would be placed in a room, and the radio would be blasted for everyone to hear. Teachers would be rendered useless as leading authorities would be the global experts in the various fields.
It didn’t happen. The idea of replacing the professionals is the same story we currently hear about AI. Will it happen? History tells us it might not.
Then came the television. It didn’t change education at all. After the TV was the compact disc. This was supposed to change the landscape. You could record videos and use them to study, preserve the lesson to replay it at one’s discretion. That’s a lot to ask from students.
It didn’t change education. CDs were merely conduits for the piracy of songs and movies.
The message is always the same. This time it’s different. This one is different. Difference, however, does not imply improvement.
AI has features of almost all the other technological advancements. Like the TV, it can create videos. Like the radio, there are also audio versions. It surfs the web and stores your chats, making it better than a compact disc.
Surely, it must be different.
But as far as base rates go, the effects would not be felt, especially if there is no effort from the individual. Education, thus, is a factory for pumping out compliant, low-agency products. The gifted individuals work to get picked rather than to pick themselves and disrupt industries and domains.
Policies not grounded and tested with the ultimate professor, time, will hardly do anything different.
Bringing down education will face significant backlash. Education, for instance, is one of the leading causes of reduced early pregnancies. Why would I want to risk such a great return for the girl child?
Regardless, if we can invent education, we can develop ways of preventing early pregnancies. We don’t have to rely on an inconsiderate processing factory to achieve this result.
What I’m trying to say is…
In medical school, there’s a saying that goes:
See one, do one, teach one.
It’s best implemented with a tutor to guide you. It’s how I performed my first CS surgery. It’s how I learned to intubate patients and insert CVCs and dialysis catheters. I had excellent tutors.
YouTube videos lacked a guiding hand. It was because I had excellent tutors that I crossed the Rubicon. It was either I learn or the patients die. So I applied effort.
We can develop such a framework, pair-wise frameworks. The sooner we do, the better for everyone since it’s not a zero-sum game.
This song inspired some of the lines used in this article. Source — YouTube