The 2 Sigma Problem

In 1984, educational psychologist Benjamin Bloom published a paper called “The 2 Sigma Problem.”

The paper highlights an important discovery. Students tutored 1-on-1 using mastery based learning performed two standard deviations higher than students who learned through typical methods. Put simply in Bloom’s words, “the average tutored student was above 98% of the students in the control class.” Further, Bloom and his research team found consistency in results from students in all grades and in all environments.

This tells us a few important things:

1) Our system is setting up students to be average.
2) All students are capable of mastery.
3) Students need personal instruction to thrive.

Let’s dig into each of these three ideas.

Our current education system sets students up to be average. It’s diametrically opposed to what Bloom found to be effective. Rather than using mastery based learning and 1-to-1 tutoring, our system consists of massive class sizes that spread teachers thin and awful grading standards that do nothing to make sure students actually learn and retain knowledge.

This is a massive problem.

Students fall out of love with learning and natural curiosity at an early age thanks to a system that better resembles quasi-educational daycares than effective learning and curiosity incubators. We can do better.

In 1968 Bloom proposed “learning for mastery” - a method where students must demonstrate a given level of mastery on a part of a subject before moving onto the next part. Rather than giving all students the same amount of time to learn a subject, each student gets as much time as they need to become proficient in the material. Our current model resembles memorization and assessment - not learning and teaching - and mastery based learning can help fix that.

Check out these two TED talks by Khan Academy founder Salman Khan for a great explanation and demonstration of how effective teaching for mastery is:

1) Let's teach for mastery -- not test scores.
2) Let's use video to reinvent education.

While mastery learning is a great step, we also need to address the problem of large class sizes and a lack of personal instruction.

We put both teachers and students in a horrible position. More students and less funding means that class sizes are larger than ever. Teachers try to make do with the resources they have, but they’re put into an impossible situation - 1 teacher isn’t enough for 30+ students. They’re spread too thin. And now they also have to juggle remote learning.

In video #2 above, Sal Khan talks about how video instruction helps address this problem. Recorded lessons allow students to pause, rewind, and rewatch the material allowing them to learn at a pace that’s natural to them at a time that’s convenient for them. Teachers can effectively record their lessons so that students can get a more personalized experience despite the lack of direct attention from the teacher. Again, this is a step forward, but true 1-on-1 tutoring requires interaction.

At the time, Bloom concluded that 1-on-1 tutoring was "too costly for most societies to bear on a large scale." He raises an important point. Getting every student a personal tutor is a burden than society can’t bear. We have neither the number of teachers needed nor the resources to fund them. However, technology will be able to solve this problem, and it will be able to solve it much sooner than we think.

Advances in AI and AR/VR technology will soon allow us to create virtual tutors for our students. Not only will this eliminate the need for human tutors, but these tutors will be far more effective than human tutors ever could be.

We’re fast approaching a world where every student will have a customizable, personalized tutor that is an expert in every subject in their pocket.

A world that looks like this:

This is the problem I’m working on at LearnFromAnyone. In front of us lies the path to building the scalable solution to Bloom’s 2 Sigma Problem, and LFA aspires to be the solution to that problem.

While implementing one of either mastery based learning or 1-on-1 tutoring will raise students by one standard deviation (84th percentile), implementing both of them will raise them by two standard deviations (98th percentile).

A future where tomorrow’s average student is equivalent to today’s 98th percentile student is a future worth building and a future worth aspiring to.