Thursday, January 29, 2015

Automatic Systematic Habit: Automaticity

Welcome, dear reader! Let's play a quick game. Go through the list below and name the color of the ink used for each word. 


Easy, right? I thought so! Let's play another game that's a little more interesting. Same drill: Name the color of the ink used for each word. Ready? Set. Go!


If you're like me, that was a little more challenging. I was a lot slower, and I made a bunch more errors. Why was naming the ink colors so hard?!


"Yeah, you're so good at it. A systematic habit" --Garbage

Before we answer why the second task was harder, let's give the phenomenon a name. In the psychological literature, it is called the Stroop Effect, named after the scientist who made the discovery famous [1].

Remember my claim that we are serial processors, meaning that we can only process one stream of information at a time? Well, that is mostly true. If we get really, really good at a skill, then it can become automatic. Once a skill becomes automatic, we can carry it out without devoting much attention to it. The process of executing a skill without requiring attention is called Automaticity.

Since you elected to click on a link directing you to a highly verbal resource (i.e., this blog), I assume you are an expert reader. Expertise, obviously, is a matter of degree, but if you've been practicing the skill of reading for more than 10 years, you are most likely an expert reader [2]. Being an expert, you have automatized the process of reading. When symbols that resemble letters are arranged in what looks like words, your brain will automatically try to read what the letters say. I also assume that, if you are older than a toddler, you have probably automatized the skill of recognizing primary colors.


Learning How to Walk, Talk, and Chew Gum

How does one automate a skill? The answer is pretty straightforward. There's no magic formula. You have to practice the skill. A lot. And preferably with implicit and/or explicit feedback. By "explicit" feedback, I am referring to a teacher or tutor who tells you that you are doing well or that you made a mistake. By "implicit" feedback, I mean the external environment signals that your performance was flawed in some way. For example, if you swing a bat at a ball and miss, the fact that you missed will become obvious to you when the ball continues past you instead of connecting with your bat.

The outcome of automatizing a skill is glorious. It means that you can carry out the task without much conscious awareness. You can suddenly (by "suddenly", I mean after 10 years of deliberate practice) do two things at once. You can finally multitask!

The downside, however, is when two automatic skills conflict. The Stroop Effect is a powerful example of this collision. The automatic skill of reading comes straight into conflict with the automatic skill of recognizing colors. When the output from each of those processes are consistent (e.g., RED), then performance is fast and error free. However, when they are in conflict (e.g., RED), then performance slows down and is fraught with errors.


The STEM Connection

Automaticity and education have a very long history together. Where did the idea of "flashcards" come from if not from the application of automaticity to education? Some skills just need to be automatized (e.g., addition, subtraction, multiplication, and division for integers up to 12). The advantage of automatizing certain skills is that it frees up the student's attention and working memory so she can devote additional cognitive resources to acquiring more complex skills. 

The problem with undergoing the process of automaticity is that it can be extremely boring. Trial after trial. Flashcard after flashcard. When will it end?! The instructional designer's job is to help motivate the learner so she can reap the benefits afforded by automaticity. Here is where I think educational games can make a positive contribution. The video game industry has figured out what motivates us. How can we take those lessons and apply them to the development of instructional experiences that help reduce the monotony of working towards automaticity?

One of my favorite examples of an educational game is Akira, which is a game that helps build fluency in decimal and fractional comparisons [3]. The goal of this game is to make number-magnitude estimates as quickly as possible. One of the important elements of these types of games is "time pressure" so that students have to work as quickly as possible, all the while avoiding errors. I also like Battleship Numberline, another educational game that attempts to build number fluency.

With the rise of mobile devices and the availability of inexpensive development tools, the potential for creating educational games has never been better.


Share and Enjoy! 

Dr. Bob


For More Information


[1] According to wikipedia, the finding had been discovered before Stroop, but he gets the credit with his publication: Stroop, J. R. (1935). Studies of interference in serial verbal reactions. Journal of experimental psychology, 18(6), 643.

[2] Ericsson, K. A., Krampe, R. T., & Tesch Roemer, C. (1993). The role of deliberate practice in the acquisition of expert performance. Psychological Review, 100(3), 363–406.

[3] Just to be on the up-and-up, Akira is currently being developed by the company I work for; however, I am not a member of that particular development team nor do I derive financial benefit from its success.

Thursday, January 22, 2015

The Double-Wide: Dual-Coding Theory

Here's your mission, should you choose to accept it. Your goal is to memorize everything that you see in the scene that is linked below. Click on the link, give yourself about 60 seconds to scan the image, then hit the "Back" button on your browser and list as many items as you can.


=======================================
=======================================

What do you remember from the scene? Were there details that were easier for you to remember than others? Did you use any particular strategies to help you encode or recall as many items as possible? What you ultimately remembered likely depends on the strategies you used to encode the scene, or, in other words, the mental representation you used during the encoding process.


Dual-Coding is like a Double-Wide Trailer: Twice the space, Twice the Fun!

So far, we have talked about a couple of different types of mental representations: 


One property that’s common to all representations is that they are incomplete. This is precisely why representations are useful. Instead of encoding every bit of possible information, representations gloss over unimportant details. The reason for this is computational efficiency. If you had to consider every unique piece of information while running a mental model, then the simulation would go on forever. 

Another property of representations is that some are better suited for specific tasks than others. Some tasks, perhaps like the one above in which you are trying to remember details of a visual image, are easier when we envision a mental picture. For other tasks, however, such as remembering items in a shopping list, it might be more useful to rely on verbal labels, either in written or acoustic form [1].

In many cases, it is possible to use more than one type of representation for the same thing. Concrete nouns, like ocean for example, can be represented in at least three ways:

  1. Orthographic Representation: a string of printed characters (e.g., in English: ocean or in Japanese: );
  2. Phonologic Representation: a sound (e.g., in English: |ˈōSHən| or in Japanese: |ooME|);
  3. Imagistic Representation: an image.

In its simplest form, the dual-coding theory proposes that an individual is more likely to retrieve information from long-term memory when the memory was initially encoded with multiple representations [2]. I am more likely to retrieve facts about the concept ocean if I have multiple routes to that pieces of information in long-term memory. That’s why dual-coding seems to work. Thus, a great way to learn a new vocabulary word would be to experience all three representations simultaneously:



Of course, life is not so simple. Most of the interesting things you want to learn about are not concrete nouns. Instead, you are more likely tasked with learning abstract concepts, like equivalent ratios and linear relationships.


A STEM Example

The same strategy that works for learning new vocabulary words also applies to mathematical instruction. It’s not that we strictly need to present a picture, word, and a sound all at the same time for each new concept. Instead, we can use multiple representations that all target the same concept. 

Let's take equivalent ratios as an example. According to Brian Wilson, there are roughly "two girls for every boy" in Surf City. How might we represent that ratio with an arbitrary number of boys?  It turns out that ratios can be represented in (at least) three different ways. 

First, the ratio can be represented symbolically: "2 girls : 1 boy" or "2 girls / 1 boy"

Second, the two numbers can be can plotted against each other on a double number line, like this:

Finally, if you rotate one of the axes ninety degrees, you form a Cartesian coordinate system; thus, the set of equivalent ratios all fall along the linear function:


If all three representations are used in a lesson about equivalent ratios, students are more likely to be able to recall and apply what they learned about equivalent ratios at a later time because they have three routes to get to the same concept. 

Share and Enjoy! 

Dr. Bob


For More Information

[1] More on memorizing lists later. In a later post, we will talk about memorization strategies that are thousands of years old (e.g., the memory palace strategy).

[2] Paivio, A (1969). Mental Imagery in associative learning and memory. Psychological Review, 76(3), 241-263..

Thursday, January 15, 2015

A Distinction With a Difference: Declarative and Procedural Knowledge

Representations in Memory: Facts vs. Skills

Let's start with a pop quiz. While answering each question, introspect on what your brain is trying to do. Think about how you are attempting to generate an answer for each question:

1.  What is the capital of Montana?
2.  Solve the following equation in terms of x: 2x + 40 = 80

Do you have some answers? Excellent! Which question was more difficult? Which one took longer to answer? Which question was more factual, and which question was more skill-based? (BTW, if we stuck your head in an MRI and started scanning, we would notice that different brain areas would have become active for the two different questions.) So, what's going on? Why are there differences?

As one of my professors was fond of saying, "If you can imagine three ways the brain can do something, it does it in all ten." The same is true with mental representations. There are several different ways the brain can represent information. For example, we previously talked about the power of mental models and what they afford us. 

While extremely useful – as we saw – mental models are only one way the mind represents knowledge. The mind also stores information using other types of representations, including declarative and procedural memories. A declarative memory is something that can be stated explicitly, like a fact. "Abraham Lincoln was the 16th president of the United States of America" is an example of a declarative fact. Alternatively, a procedural memory is one that encodes how you perform an activity or skill. Knowing how to tie your shoes is a procedural memory that has a muscular component (i.e., a "motor memory"), and knowing how to do long division is an example of a cognitive, procedural skill. What's the point of having different types of memory? What are some of the properties of these different forms of memory [1]?


Properties of Declarative and Procedural Knowledge


The first property has to do with the speed by which these memories are accessed. Declarative memories are slow to access, which makes intuitive sense. Think about how difficult it is sometimes to remember a fact, especially if it's been a really long time since you've thought about that particular subject. Procedural memories, on the other hand, are very fast. You can fire off a procedural skill in very little time, sometimes almost automatically. This automaticity is often a good thing in that it lets you execute complicated skills without having to ruminate over every step required along the way. 

There is a trade off, however. The availability of each representation is differentially influenced by context. Declarative memories are generally available, regardless of the current circumstances. That is, they are independent of the context. You can recall the 16th president anywhere on Earth. Procedural memories, on the other hand, are more sensitive to the current context. For example, it might not occur to you that a specific skill is required when there is nothing in the current environment to cue that skill. Thus, we might describe procedural skills as context dependent (see Table 1).


Table 1. Properties of Procedural and Declarative Memory

Finally, I think it is important to make these distinctions because the representation knowledge changes over time. When you first start to learning something new, the knowledge is generally represented declaratively, but can become procedural over time. Take learning how to drive a car as an example of the process of declarative knowledge becoming proceduralized: 

A lot of what your driving instructor told you was verbal. You had to learn where the gas and brake pedals were. You also had to learn where the turn signal was, the headlights, and all of the various buttons and levers that are required to drive an automobile. You also had to learn a ton of traffic laws, all of which, I'm guessing, were stored as declarative chunks of information. But then, as you became an expert driver (i.e., over the next 10 years), you didn't represent any of that knowledge as explicit, declarative facts. Instead, you no longer had to think about where to place your feet, or what do when changing lanes (i.e., check your mirrors, check your blind spot, signal, etc.). It all became procedural knowledge and happened automatically. The same is true for cognitive tasks. Learning to solve routine problems can eventually become automatized, which means that the declarative representations you had in memory are now so automatic that they are converted into procedural memories [2].


A STEM Example

How does knowing the distinction between declarative and procedural knowledge help us improve education? First, it is has implications for how we train our students. When teaching a new skill, a good approach is to provide a conceptual introduction by articulating a set of declarative chunks of information. As discussed previously, we can increase the odds of storing declarative memories in long-term memory by finding a hookConnecting the current set of facts to some other piece of knowledge that our students already have gives them a better shot at remembering the new facts later. (I tried to do something similar at the beginning of this post: I assumed that you already knew what a fact and skill were, so I tried to map the concept of declarative [fact] and procedural [skill] knowledge onto those concepts.) 

Once you've motivated the lesson with a conceptual introduction and covered the declarative facts that are needed to develop a new skill, it is time to start deliberately acting on those facts until they are transformed into a skill. Here, it is important to give students plenty of opportunities to put their new-found declarative knowledge into action, accompanied by lots of feedback along the way. That feedback can come from you (as the teacher), other students, tutoring software, or pretty much any source that lets students know when they are on the right track or need a course correction. Again, the analogy to driving a car is pretty good. You need lots of hours behind the wheel, with lots of feedback (from an instructor, mom or dad, a kindly police officer, a not-so-kind driver in the other lane, the curb, the grinding of the gears, etc.) before the declarative facts of how to drive become procedural skill.  

Share and Enjoy! 

Dr. Bob


For More Information

[1] Nokes, T. J., & Ohlsson, S. (2005). Comparing multiple paths to mastery: What is learned? Cognitive Science, 29(5), 769–796.

[2] My favorite example of the automaticity of knowledge is my copy code. When I worked at LRDC, they had a couple of copy machines on each floor. Each person was given a "copy code" which charges the copies that you make to your account. While conducting a study, my research assistant asked me for my code, and I could not verbalize it! I had to "let my figures do the talking." I had to type the code and look at the numbers that I was hitting. The memory was completely converted from a declarative memory to a procedural one. 



Thursday, January 8, 2015

Have You Gone Mental?: Mental Models

Using Models to Reason and Infer New Knowledge

Let me ask you a question: How many windows are in your house or apartment? It's entirely possible that nobody's ever asked you this before. At least, that's what I'm banking on. If you've never been asked "How many windows are in your home?", then that means you aren't answering from memory. Instead, the question requires that you compute a value on the spot. How did you accomplish this task?

My prediction is that you visualized your home, and then started a walk-through, counting each window as you moved from room to room. In other words, you used a mental simulation, or a mental model, to answer my question. As it turns out, mental models are great for more than just answering random questions. They are just one instance of a class of mental representations that we use everyday. Mental models are simulations or images that we use to reason about the world and/or infer new knowledge.


What do toilets and light beams have in common? 

Consider another example of a mental model: the flushable toilet. If you know how a toilet works, then you can use your mental model to debug it when things go wrong. For example, a well-constructed mental model will help you figure out why the water keeps running (i.e., the filler float is stuck or the flush valve is stuck in the open position). Or why nothing happens when you depress the handle (i.e., the chain that connects the handle to the flush value fell off or is broken).

In addition to reasoning about the world, mental models are also useful in generating new knowledge through the process of inference. In a previous post, we talked about the power of inheritance to derive new information. This is similar in the sense that you infer new facts by "running" a mental model. 

One of the more famous examples of this is Einstein's claim that he used a mental simulation of riding a beam of light and asking all sorts of questions about what he might observe at that speed. Good thing he interrogated his mental models because it gave birth to the special theory of relativity!


A STEM Example

There are so many examples of mental models in science and engineering that I won't even attempt to catalog them here. In fact, one could argue that STEM education is primarily focused on helping students build detailed and accurate mental models. Here are a couple of illustrative examples.

First, the astute reader probably noticed that a recent post, entitled "Midnight in the Garden of Encoding and Retrieval," attempted to create a mental model of memory. That model proved to be useful when we started asking questions about what happens during encoding, storage, and retrieval. The answers to those questions helped us debug potential reasons why a student might fail to learn a new fact or skill. 

Another example, that I've used in my own research, is the human circulatory system. In one of our studies, we asked about the thickness of the muscle for the right ventricle versus the left ventricle of the heart. If you know that the right side of the heart sends the blood to the lungs, and you know that the lungs are proximal to the heart, then you know it doesn't need to pump very hard; therefore, the muscle in the walls of the right ventricle do not need to be as thick as the muscles in the left ventricle. This is useful knowledge that doesn't need to be taught directly. Instead, it can be inferred by the student through a series of leading questions. 

The final example I will give is one of my favorites [1]. It has to do with the development of the mental model for the Earth. When kids are little, they know, via observation, that the Earth is flat. Later on, they learn that the Earth is round. To make the observation compatible with the authoritative knowledge that they hear from adults, children then reason that the Earth must be round, like a pancake. If you ask them leading questions, such as, "What will happen if you walk for days and days?", they will answer that you will come to the edge of the Earth. 

If the goal of education is to help students develop accurate and complete mental models, then there is a pretty interesting implication for assessment. It is difficult and time-consuming, but developing generative questions is a excellent way to evaluate your students' mental models. Generative questions ask the student to reason about his or her model model. The "muscle thickness of the ventricles" and "walking the Earth for days and days" are good example of generative questions. 

Share and Enjoy! 

Dr. Bob


For More Information

[1] Vosniadou, S., & Brewer, W. F. (1992). Mental models of the Earth: A study of conceptual change in childhood. Cognitive Psychology, 24, 535–585.

Thursday, January 1, 2015

Midnight in the Garden of Encoding and Retrieval: Memory Models

A Simple Model of Memory

In a recent post, I made the following claim: Learning cannot occur when there is no attention. In other words, information must pass through all of the attentional filters before learning can take place. We also talked about working memory as an important buffer, and it is a location where information temporarily resides. But what happens after that? How do we store information for later use? Once we have a model for learning, then we can theorize where things might go wrong. If we better understand where things go wrong, then we can help debug those processes and help our students do a better job learning new information.

To motivate this a little bit, remember the first time you learned the Macarena? You first watched people flail on the dance floor in perfect synchrony. Then, you practiced the dance steps in the privacy of your bedroom. Finally, at the next party, you expertly threw down all of the moves. How does that map onto a simple model of memory?

Let us assume that information has gone from your sensory register (a buffer that holds a ton of information but for only a very brief duration) and through the selection mechanism employed by the attentional system. Now the information is in working memory. What next? Here is a simple process-model of memory that we can use to track what happens next.



Through the process of encoding, information is passed from working memory into long-term memory. There, the information undergoes a storage process. And now for the moment of truth. The final step in this model is retrieval, where information is pulled out of long-term memory. Retrieval is when you need to recall a fact (or procedure) and put it to use. 

Debugging the Process

Each step outlined above has some probability that it will fail. It doesn't necessarily mean something is wrong (only that we are human). Let's consider each process independently. 

Encoding: As a learner, you may fail to encode the information in a precise fashion that helps you recall it later. Let me give you a perfect example of a problem with encoding. You've seen a penny before, right? So which way is Lincoln facing? What words, if any, appear above his head? Is there any information to the left or right of Lincoln? If you're like most people, this is a very difficult task. It's even hard when you are asked to pick the real penny from a lineup of fakes, as opposed to recalling all of the various features. If you struggled with this, it's because you never bothered to encode the features of a penny. And why should you? They are readily available, and it is likely that nothing really depends on you encoding all of the features.

Storage: There is some decay function associated with memories held in long-term memory. Memories typically decay when they aren't actively used. There are exceptions, of course (i.e., permastore [1] or flash-bulb memories). But for the most part, memories that aren't used fade from long-term memory. Think back to your first history class. If you're like me, it's been a while since you've thought about the names of the U.S. presidents, in order, and the dates they were in office. If so, then it's likely those memories have faded away. 

Retrieval: Finally, there may be a problem when you go to remember something. You know it is there, but you can't get access to it. A great example of this is called the "tip of the tongue" phenomenon. It's largely a problem with retrieval because you know you know something. But at the time of retrieval, something is blocking your path to that information. Often, the memory is blocked because of interference from some other, related (but irrelevant) memory. A good strategy for getting around problems with retrieval is to leave it alone, and try recalling it at some later time. The reason this works is because the interfering memory has started to fade away. 


A STEM Example

Suppose you are tasked with learning a new procedure, such as programing a computer to add two integers. Your background research suggests Python is a great programming language for beginners because the syntax is simple. You also are delighted to discover that your computer already has Python installed. 

To start the interpreter, you locate and launch the application called Terminal. Then you type python to start a session. Next you learn that a program is a collection of functions, which are small blocks of code that do something useful. To define a function, you use the keyword def followed by the name of the function, any arguments you wish to include in parentheses, followed by the colon character ":". All functions require a return value, which you specify. Your little program ends up looking like this: 

def add_two_numbers(addend1, addend2):
    sum = addend1 + addend2
    return sum

By my count, putting this together requires learning at least 8 different pieces of new information. Some of it is pretty arbitrary (e.g., using a colon to close the definition portion of the function), and some is highly conceptual in nature (e.g., a program is a collection of functions).

Now, suppose you want to teach the above lesson to someone, but you soon discover that his or her program does not work. To help diagnose why, you might ask a series of questions to figure out which memory process is at fault. Is there a problem where the information was never encoded in the first place? Has there been a long lag between the initial encoding and the first attempt to retrieve the information? If so, then there might be a problem with decay. Those memories might just have faded because of the lack of use. Finally, it could be a problem with retrieval because Java is kind of like C++, which may share some syntax with Python.

I admit, when I wrote this up, I had to go find an old Python program that I wrote a couple years ago because I couldn't remember which keyword to use. Namely, I was getting interference from Lisp, which uses defun instead. It is likely that I never actually encoded def because I always have some example code laying around (like pennies).

The model presented above is an overly simplistic version of memory. But like most things in life, problems seem simpler to solve when you have a nice working model for how they should operate in an ideal setting. 

Share and Enjoy! 

Dr. Bob


For More Information

[1] The idea of a permastore is extremely interesting, and probably warrants a separate blog in and of itself. Basically, it's the hypothesis that there are some memories that you create that will never fade away, no matter how old you get. The hard part, of course, is verifying the veracity of those memories, especially if they are auto-biographical.