Education

I believe that schools can be step-change improved, but “how we teach” probably doesn’t need to change as much as many radical movements think.

Disclaimer: I’m obviously not a pedagogical scientist or teaching professional, so consider this a set of believes I hold rather than an expertise I claim to have.

Curriculum §

Curriculum, honestly, should be easy. There’s plenty of good science around how to teach specific understanding.

Direct instruction works. There is evidence that concreteness fading works in maths, and I believe that should transfer (probably with nuance) to other subjects also. There is much strong support for spaced repetition. Additionally, integrating skills in cross-curricular learning is supported in STEM but needs more nuanced research for other fields¹.

Specifically for language learning² we know that explicit instruction does really well for beginning learners, while immersion learning realizes much better long-term retention and natural fluency. As such, I believe³ it should be taught explicitly quickly and briefly at a young age for fast understanding, after which it can be taught implicitly in other classes.

Using these basics, a single integrated curriculum across subjects can be designed. Note that these effects and their applications differ for age groups, and should be applied differently in the curriculum.

Teachers can be instructed very specifically to teach this curriculum, which should improve their efficacy and reduce the impact of potential teacher quality differences.

Teaching methods §

Mastery learning is strongly supported by evidence. Especially having stronger effects on weaker students, who need more attention. The curriculum should be trackable (I imagine a knowledge-tree here), and teachers should collaborate deeply and often to understand where students are at and what they need to gain mastery. Velocity through branches can be used to identify where students need additional support. Ideally this can compensate for some external impacts, like parents.

Peer tutoring as well has much supportive science, and can be used to reduce teacher time use while benefitting student outcomes as well as behavior in schools at the same time.

School systems §

(todo)

• consistency is hard
  • teachers that want to do the same all the time
    • older students help out younger students
    • part of curriculum!
  • or at least feel autonomy and impact
    • alumni involvement
    • focus on goals and simple methodology
    • obfuscate individual teacher impact?
      • what’s the value of having the same teacher forever?
      • is it always better, how often is it worse? Is that worth it?
    • if the system is good, teachers can fall back on it (ie be worse)

Footnotes §

1. This could also free up resources. Also, just imagine the possibilities!

   • statistics through elections
   • calculus through budgeting, taxes
   • reasoning through civics, history
   • reading and other media for bias recognition, fact checking, etc
   ↩

2. Oops, a bias for one subject. Language learning is of special interest to me, as I’ve experienced both explicit immersive and explicit instruction methods (and I suppose implicit immersive through natural, self-motivated exposure learning) and have strong feelings about the efficacy of either. ↩

3. I also believe this should then be researched and validated, of course… Assume that for all systems from here on! ↩