Saturday, January 17, 2015

Make It Stick - The Science of Successful Learning

While I've been sick, I listened to a very interesting audiobook: Make It Stick - The Science of Successful Learning, by Peter Brown, Henry Roediger III, and Mark McDaniel.  It currently has 4.5 stars on Amazon from 151 reviewers.  This is not the Heath brothers' book Made To Stick, which is a good book about how to convey your message so it will stay with your audience longer.  I wrote about that one back in 2007.

Make It Stick is about learning.  How people learn.  What works.  What doesn't work.  And why.  It throws cold water on the standard practices of re-reading material, study groups, and some routine sports drills.  And it describes methods that have been shown to work much better for learning and long term retention.

This book is full of helpful tips and guidance for anybody wanting to learn or teach.  That's basically everybody.

Below I've listed some of the concepts presented in the book.  If any of these intrigue you, then you'll probably enjoy the book.  I apologize if I misunderstood any of these concepts... I was sick.

Effortful learning lasts longer
  As it is with lottery winnings, a pile of knowledge handed to you doesn't tend to last very long.  We retain what we learned better and longer if we had to use some effort to learn it.

Rereading and highliting text is not very effective
  Re-reading material can result in self-deception: we think it works because we become more familiar with the material.  This can be helpful if cramming, but it will not last.  Re-reading does not result in durable memory.  It is time consuming; time that is better spent on other methods.

Familiarity is not mastery
  This is a recurring theme in the book.  Many popular learning techniques are good at building familiarity.  We often confuse this with learning.  While familiarity is definitely useful, it is not the same as learning the subject matter.  To trivialize and over-simplify this to an extreme: An average American can live in Paris for 3 months and they will gain familiarity with the language.  However if you ask them to write a 3 page essay on any topic in French, you will be disappointed.

Retrieval practice produces better learning
  When we practice remembering what we have recently learned, we will remember more and it will last longer.  Flash cards are a simple example of this method (although it's important to shuffle the deck and to not remove cards too soon).

Testing and quizing is very effective, even if it doesn't count
  Tests are effective because they exercise the retrieval method.  They also help us to identify our weaknesses, and to focus future learning activity on those weaknesses.  Teachers and college professors have started to notice this: giving frequent quizes and tests consistently results in better overall scores and grades, even if the quiz scores are not counted (and the students know they don't count).

  Learning is improved when we revisit material at a later time.  This exercises the mind to recall the material and reinforces the memory.  Spacing is far better than burning something into memory through sheer repetition.  So, spending 100 hours in one go to drill yourself on a topic is less effective than spending 100 hours spread over a period of time.  It helps most if the time gap is just enough that you're beginning to forget.  Then retrieving the memories of what you previous learned helps to strengthen the neural synapses and build new ones.

  We're all famililar with the method of practicing one thing until you get really good at it, then you move on to the next thing.  This is called "mass practice" and it's quite effective.  The book describes how mass practice is good at short term memory but is not ideal for long term performance.  You can improve your overall training performance if you interleave practice of one thing with practice of another thing.  This break in the linearity helps us to solidify our learning and it will last longer.

Difficulties help us learn
  Airline pilots spend valuable time in simulators going over emergency situations - challenges - difficulties.  This is by design and it is very effective.  The SEALs have a saying that goes something like: the more you sweat in training, the less you bleed in combat.  There are tons of similar examples where people choose to employ difficulties in their training regimine in order to boost their overall performance.

Generation is more effective
        Filing in the blank is more effective than picking from choices.  Multiple-choice tests are fun because it's relatively easy to achieve the level of familiarity needed to perform well on them.  Recognizing something is not the same as knowing something.  So, if you're going to design a quiz (for others or for yourself), then use blanks and not multiple-choice.  I think "generation" is a variant of "retrieval" in the book's vernacular.  But I'm probably missing some subtle distinction.

No comments: