Dave's Whiteboard

It’s a new week.  For some people, it’s a new year.  The equinox, as Wodehouse said of Christmas, is at our throats again.  What better way to clear your brain than considering the pentatonic scale?

(Note:  I’ve switched the video to the version on YouTube; the original one seemed to perform poorly when embedded here.  You can find that original,  Bobby McFerrin Demonstrates the Power of the Pentatonic Scale at the World Science Festival on Vimeo.)

I’ll bet a pretest would have indicated that not many of these people could read music and that not many of them are prone to sing in public.  Pay attention to how little McFerrin instructs and how much he proposes.

I wasn’t much surprised to find that one panelist is Daniel Levitin, author of This is Your Brain on Music (mentioned a time or two here on the Whiteboard).  You’ll find the entire Notes and Neurons discussion in five videos (lengths vary; about 75 minutes in all) at the World Science Festival website.

Is our response to music hard-wired or culturally determined? Is the reaction to rhythm and melody universal or influenced by environment? Join host John Schaefer, Jamshed Barucha, scientist Daniel Levitin, Professor Lawrence Parsons and musical artist Bobby McFerrin for live performances and cross cultural demonstrations to illustrate music’s note-worthy interaction with the brain and our emotions.

What more invitation do you need?

Yesterday’s New York Times ran an article about the effect technology can have on families (Breakfast Can Wait.  The Day’s First Stop is Online).

You can imagine the examples: “All four of us starting the day on four computers in four separate rooms,” says one parent.  A fourteen-year-old “went from walking the family Labradoodle for 20 minutes each morning to only briefly letting the dog outside” — and blames Facebook (and the peer connections it delivers).

The Times article seems an apt companion to a New Yorker article: Don’t! (subtitle: the secret of self-control).  Based on experiments he conducted at Stanford with young children, Walter Mischel believes a child’s ability to postpone eating a marshmallow was a good predictor of things like the ability to plan, maintain friendships, and do well in school.

Marshmallows?

The experiment involved nursery school children.  A child picked a treat (like a marshmallow or a cookie).  The experimenter made an offer: you can eat the one treat now, or, if you wait while I’m gone, you can have two when I return.

Once the experimenter left, the child could also ring a bell, which would bring the experimenter back.  (My guess is that this allowed the child to put an end to what must have felt like endless waiting.)

What happened?

[The typical child] struggled to resist the treat and held out for an average of less than threde minutes….About thirty per cent…successfully delayed gratification til the researcher returned, some fifteen minutes later.

Mischel’s conclusion is that the crucial skill for these “delayers” was ” the strategic allocation of attention.”  They found ways to distract themselves from the treat.

“What’s interesting about four-year-olds is that they’re just figuring out the rules of thinking,” Mischel says.  “The kids who couldn’t delay would often have the rules backwards.  They would think that the bst way to resist the marshmallow is to stare right at it, to keep a close eye on the goal.  But that’s a terrible idea.  If you do that, you’re going to ring the bell before I leave the room.”

Mischel and others are currently on a new study on whether self-control skills can be taught — a form of metacognition, a way to help children manage their own behavior.  Last year, for the charter-school organization KIPP, Mischel and Angela Lee Duckworth are currently working at this.  They find significant improvements in things like the ability to deal with “hot emotional states,” though they’re cautious about claiming any long-term results yet.

This TED talk by Joachim de Posada deals with those Stanford experiments.  As he says, “To tell a four-year-old kind to wait 15 minutes for something they like is equivalent to tell us, ‘We’ll bring you coffee in two hours.”

I can imagine a whole trademarked line of “Don’t Eat the Marshmallow” stuff, and in fact there are books and seminars.  What I’m more interested in is how solid the connection may be, and how people can build their ability to delay.

My image adapted from two CC-licensed photos:
A stopwatch by casey.marshall and marshmallows by John-Morgan.

You know what the difference is between good habits and bad ones?  Good habits, you have to work at.

In other words, they require discipline, which means willpower.  In a recent post at the SharpBrains blog,  Daniel Goleman offers good news and bad news about willpower.

The bad news is that research suggests we have a fixed amount of it.  Apply it to one challenge, and you have less available for the next.

This may explain why a (reasonably) vigorous workout for me is often followed, some hours later, by a fuller meal than I might actually need.

The good news, says Goleman, is that we can increase that reservoir of willpower.  He compares it to a muscle, which gets stronger with exercise.

(If you like Goleman’s post, you’ll probably like Sharon Begley’s book Train Your Mind, Change Your Brain. As I noted here some time ago, the title sounds like a self-help nostrum, but Begley in fact assembles and makes clear a wealth of research-based understanding about the brain.)

CC-licensed daily-discipline photo by blue out.

Series: Ten Steps to Complex Learning

Note: this is a continuation of the previous post in this series,
because I can’t seem to summarize and comment on one of the
Ten Steps to Complex Learning in a single post.)

Step 3 is “set performance objectives.”  As the introduction and first section of this chapter emphasize, this is an iterative process, not a linear one.  The real-life tasks in which you perform the complex skill help to determine the overall learning goals and the specific tasks that will help achieve them.

In turn, these help refine understanding of the complex skill and the constituent skills that it embodies.

After analyzing (or “decomposing”) the skills, you create performance objectives.  I’ve discussed Van Merriënboer and Kirscher’s actions and the tools and conditions that apply to the objectives.  It’s a bit tough to talk about standards as they describe them.

Keeping to the standards

The Ten Steps sees standards as having four elements: criteria, values, and attitudes.  Criteria means what you think: minimum requirements for things like accuracy, speed, quality, and so forth.

Values indicate that the constituent skill conforms to some set of rules or regulations.  Two examples vM&K offer: “without violating traffic rules” and “taking the ICAO safety regulations into account.”

My feelings are mixed.  I can see the value of this as shorthand (“wiring for this remodeling must meet the National Electrical Code”).  Is there a little game of gotcha on the side?  Or are we acknowledging that in complex learning, there are areas of performance that matter, even if we’re not going to provide instruction related to them?

I really can’t say; this just feels a bit like a junk drawer in the conceptual cabinet of the Ten Steps.

Feelings about attitudes

If values are the junk drawer of instructional design, attitudes are like scribbling “Get organized!” on a to-do list.  The Ten Steps doesn’t define “attitudes,” but says they’re “subordinate to, but fully integrated with” constituent skills.

Apparently we’ll know them when we see them.  However, they won’t be things like “have a client-centered attitude.”  vM&K say this is a non-example: a research librarian doesn’t need to have such an attitude outside of work, nor does he need to have it when doing tasks that don’t involve clients.

So, whatever an attitude is, it’s not an enduring part of your personality.  I actually think there is such a thing as attitude; I just don’t think  you can influence it directly very well.  The Ten Steps seems to agree:

It is only necessary to specify the attitude in the performance objective for these relevant constituent skills. If possible, all observable behaviors that indicate or demonstrate the attitude should be formulated or specified in a way that they are observable!  The standard “with a smile on your face” is more concrete, and thus more observable, than “friendly;” “regularly performing checks” is more concrete…than “punctual…”

“With a smile on your face?”

This is an unsatisfactory nod toward a complex issue.  Think of medical professionals interacting with patients (so-called bedside manner).  Can it be that helping a surgeon demonstrate interest in the person and not just the condition–”Dr. Manoogian, your gall bladder’s in room 5″–might require this close a focus?

Classifying Objectives

Three dimensions apply to the objectives you develop (remember, these are objectives for the constituent skills that make up the overall complex skill):

• Teach, or not?
• Non-recurrent, or recurrent?
• Make automatic, or not?

The easy part is sorting out the objectives you’re not going to include in your training–either because the typical performer already has these skills, or because the objectives are covered elsewhere.  Those that remain fall into four groups.

Non-recurrent skills, you’ll recall, are applied differently to different problem situations.  They involve schema-based problem solving and reasoning.  They require supporting information like cognitive maps during the training, which is the topic for Step 4.

Recurrent skills are those you apply the same way to different problem situations.  They’re the rule-based skills.  In training, these require procedurall information.

vM&K state that any prerequisite skills for a recurrent skill are by definition recurrent.  “A recurrent constituent skill,” they says, ” can never have non-recurrent aspects!”  Since they say with with both italics and the second exclamation point in two pages, they must mean it.

The same skill, they go on, could be non-recurrent in one training program, but recurrent in another.  Repair of military aircraft in peacetime might be a non-recurrent skill, because there’s time for diagnosis.  In wartime, one of the criteria is to repair or replace as quickly as possible, which could mean that repair becomes a more procedure-driven and thus recurrent task.

Some recurrent skills require a high level of automaticity.  This involves the part-task practice discussed in Step 10.  Some jobs don’t require this type of automaticity (for example, the recurrent patent-examiner task).  Factors that suggest automaticity include:

• Enabling other skills higher in the hierarchy. Musicians practice scales, even after achieving a high level of skill, in order to automate basic skills and enable more fluid performance of higher skill.
• Simultaneous performance with many other skills. Process operators in manufacturing and air-traffic controllers are two types of jobs where the individual reads displays automatically as she analyzes and responds to dynamic sistuations.
• High risk in terms of cost, damage to equipment, or danger to life. Pilots and flight attendants practice emergency procedures.

Ten Steps makes a point that not all routine skills need automation.  There’s a cost/benefit consideration — you don’t memorize all the addition possibilities of two numbers from 0 to 999; you do (eventually) automate the skill needed to keep a car moving in a straight line.

Twofers

“Rare situations” exist, according to vM&K, when you’d choose to automate a non-recurrent skill.  They use “double classified” for what seems to be combinations of recurrent and non-recurrent skills, like their example of shutting down a power plant in an emergency.

The shutdown can occur in many ways, depending on circumstances (non-recurrent), but must follow specific procedures (recurrent).  This is an expensive decision and often requires high-fidelity simulation.  In addition, the authors say that learners should be explicitly told that they’ll switch from automated mode to problem-solving mode at times.

The things you left out

Remember that “category” of objectives that you won’t be teaching?

If learners have already mastered a particular constituent skill in an isolated manner, this is no guarantee that they can carry it out in the context of whole-task performance.  Performing a particular constituent skill in isolation is completely different from performing it in the context of a whole task, and automaticity of a constituent skill that has been developed through extensive aprt-task practice is often not preserved in the context of whole-task performance.

Which is to say that when Bruno gets a perfect score on the loan-application system, it doesn’t necessarily mean he can use it while interviewing a live loan applicant at the bank branch.

Objectives and assessment

It’s pretty obvious that clear, observable objectives relating to clusters of constituent skills that make up the complex skill have many benefits.  You can develop tools to help learners do self-assessment.  You can provide support for a peer, who can help identify areas of improvement (and whose own performance can benefit from helping the other person).

IN assessment, values and attitudes are usually measured narratively, or through qualitative scales (very poor, poor, acceptable, good, excellent).

vM&K acknowledge the potential burden of a highly-detailed assessment, which is virtually a necessity for complex sills.  They recommend self-assessment and peer assessment.  They also suggest a development portfolio, a collection of assessments for all learning tasks.

This details what the learner’s done and how well he’s doen.  He can choose his next learning tasks based on this information.

In the next post, we’ll (finally) move from the learning task component to the supportive information component.  The corresponding steps are designing supporting information (Step 4), analyzing cognitive strategies (Step 5), and analyzing mental models (Step 6).

A pair of interesting posts from Cognitive Daily:

• Is expert memory better than non-expert memory?
• What is a visual expert?

(The first post includes a mini-experiment for you to participate in.  Give it a try.)

Dave Munger quotes a study suggesting that experts have a larger visual working memory in their area of expertise than non-experts do.  But he cites a second study suggesting that experts don’t have a larger working memory; their memory is more detailed.  Think of the difference between having a larger computer, and having a computer with many different applications on it.

Munger’s second post explores whether those experts are good at recognizing objects other than faces–since humans are really good at face-recognition.

Again in the second post, there’s a quick experiment for you to try.

The conclusion Munger draws from the second experiment is that “for visual short-term memory, expertise isn’t about prior knowledge, but the ability to process visual images more efficiently.

So what? In part, this suggests that expertise is not (surprise, surprise) an inborn trait. You don’t get to “expert” level because you had more parking space in your brain; your work increases you ability to make use of the space you do have.

It’s not so much knowing a lot of things; it’s being able to recognize and relate patterns.  We know a lot about ways to help people purposefully increase their ability.  Not everyone can end up as a world-class expert, but it’s certainly possible to shift the average performance upward.

(Added 3/3/09: it’s significant that I posted this four days ago and only now noticed I hadn’t included a title…)

Gadget-pyramid image by andyi, used under a CC license.