Hacker News new | ask | show | jobs
by jacobolus 3365 days ago
Who ever said there was a “sinister plot” involved?

The best (most time efficient) way to learn factual knowledge is using some kind of spaced repetition.

The best (most time efficient for the student) way to learn mechanical or conceptual tools to mastery is to be presented with a problem which needs the tool, struggle with it for a while and develop some personal familiarity with the problem and personal ad-hoc methods; then have the standard tool presented, then go back and struggle some more using the standard tool to solve various types of problems and combinatorially trying the new tool out with various old tools to see what it can do and how they all relate. Then after basic competence is achieved, continuing to occasionally apply the tool to new problems over the course of months or years, including problems which require unusual uses of the tool or unexpected combinations with other tools, generalizations of the tool, etc. This is all generally much faster and more effective with some amount of 1-on-1 tutoring/mentorship between the student and a subject expert than with any other known arrangement, whether the tool in question is a pair of scissors, argument analysis, google search, rhetoric, regular expressions, or complex multiplication.

The best way to learn methods of inquiry or creative work which characterize various human endeavors (history, journalism, physics, graphic design, ethnography, mechanical engineering, pure mathematics, philosophy, literary criticism, acting, technical illustration, sales, ....) is to spend time actually doing real or realistic work, watching experts do real work, and looking at lots of finished work.

Unfortunately it’s hard to make 1-on-1 tutoring or research project mentorship scale to a classroom. Strategies like putting the students into groups to solve project-type problems of a variety of scopes, “flipping” the classroom by presenting lectures via video outside of class, mixing up assignments to include a mix of easy small-scope problems of the new material combined with harder problems highlighting past material, and so on can help, but there is a huge gap still between (even average quality) 1-on-1 tutoring and the best alternative method, which educators and education researchers would love to know how to fill.