kitchen table math, the sequel: what is skill?

Wednesday, January 27, 2010

what is skill?

Skill is a level of performance in any given task that can be acquired only through practise. Indeed, one can consider any skilled professional as a person who has had the motivation to practise one thing far more (for approximately 10,000 hours extended over more than 10 years20) than most people could endure (BOXES 2,3). Across a wide range of tasks, the relationship between one measure of skill, the speed of task completion and the number of practise trials is well approximated by a power law21 (FIG. 1). This implies that performance continues to improve with task-relevant practise indefinitely, although the rate of improvement declines over time. Of course, most of the relevant data comes from tasks learnt for short periods of time in the laboratory. However, it is worth highlighting one classic study that reported performance of an industrial cigar rolling task22. The study included workers who had produced in excess of ten million cigars over seven years of work and they were still getting faster!

Inside the brain of an elite athlete: the neural processes that support high achievement in sports
Kielan Yarrow, Peter Brown and John W. Krakauer §
Nature Reviews Neuroscience Volume 10 | AugusT 2009
p. 587


ElizabethB said...

Cool, I like it!

I would think there there would be some sort of upper limit for cigar rolling...evidently not!

I like to use sports analogies for my remedial reading students, here's a quote from my phonics and football page:

Reading is like football. Both are complex processes built on basic skills. Anti-phonics voices lament the drill-and-kill and repetition of phonics. However, a good football player knows the importance of the basic drills and their repetition.

Even pro football players run hundreds of drills to improve their skills. While they also play practice football games, they know that you improve through mastery of basic skills learned in rote drills. Whole word methods or whole language methods are like playing hundreds of practice football games without learning how to throw the ball or run a pass pattern.

That's actually most of the page, but here's the link:

SteveH said...

Is skill just about speed or is there something else going on. When educators talk about "drill and kill", they are avoiding the issue of experience and only looking at a motivational aspect of learning. However, this requires them to greatly diminish the importance of speed and skill mastery.

There is a linkage between mastery and understanding. The problem is that this is not necessarily just about speed. It's about mastery. I don't really care if a student can do long division really fast by hand (the problem is that many educators don't like long division at all), but I do care that they can manipulate fractions very fast. This is not about speed using a rote algorithm. It's speed that comes from understanding and practice. Practice develops understanding and understanding develops speed.

What many educators seem to do is to extrapolate learning simple algorithmic skills like long division to more complex and abstract tasks. They begin to see practice and speed and mastery as less important than some sort of conceptual understanding. With a proper curriculum, speed is an indication of mastery and understanding. There may be points of diminishing returns, but the relationship is still there.

My son runs into problems in algebra that slow him down because he gets confused by a problem that is set up differently. The solution is not some sort of flexible thinking process, but the need for more mastery of algebra.

Many educators seem to view math as some sort of Zen-like thinking process that is not governed by specific skills and algorithms that have to be mastered.

Anonymous said...

"Is skill just about speed or is there something else going on."

There *can* be more going on. I suspect that the researcher's like speed because it is nicely quantifiable and also fairly easy to measure.

To support my contention that there can be more to skill than speed ... let's consider chess. An international master can make better moves than a mid-level amateur player. This is true even if the amateur is given *much* more time than the master. The master level player is simply capable of finding/selecting moves (often quickly) that the amateur just won't find.

Having said this, my guess is that one aspect to mastery is that you *CAN* get things done faster (or maybe that on average, a group with mastery will get things done faster than a group without).

-Mark Roulo

Allison said...

The chess example is excellent.

Skill means you've learned something so well that you no longer need to think about how you do it; in fact, you probably can't even articulate what you are doing.

It's brain chunking: a whole set of little skills are chunked together and repackaged in the brain; they aren't thought of as separate tasks, but entire modules. and similar to software or hardware, those modules are then essentially precompiled and stored for use later; sometimes, the hardware level itself in the brain is adapted to those chunks, just like special processors on a board that do, say, matrix multiplication.

In chess, this means skilled players can look at a board and know implicitly that certain strategies/gambits/sets of moves n ahead are available without needing to think through each node of the decision tree: they've precompiled all of that, so to speak.

Airline pilots are often so adept at flying that they can't even tell you what it is they actually do--if they try, they get it wrong. This is true in driving or any other skill as well. We learn so well that we "forget" so to speak. We no longer need to think about the surface layers, as all of those layers are done at a non/sub conscious level and we aren't even aware of them. We are able to concentrate on other elements of the task.