Sunday, May 30, 2010

Taz Talks -- Successive Approximation

OK, frankly this blog has gotten a little too grim and preachy for my taste. I think it's high time that we got back to a subject of interest. In other words, how to train a human. There is a very important concept in training that we have not talked about, and that is called Successive Approximation. In normal people language that means, "I will help you figure out the game by rewarding you for each step in the direction of what I want." For instance, when Mama taught me to jump in the laundry basket she rewarded me for coming near it, then for pawing it, then for putting my head in, then for putting one front paw in, then for two front paws in, then for hopping all the way in and then out again, and finally for jumping in and staying there.

Let us take a more interesting example. (Who wants to sit in a laundry basket, anyway? Actually, the cat does. We had to take the cat out so that I could play the game. Although it might have been more funner to jump on the cat.) Anyways, say you want to get a person to scratch your butt. Well, you could just back on up to them, which often works, but what if they don't get it, or they say "Bug off, Tazzie, I'm trying to work here."? Then you need to train them step by step.

Start by putting your head in their lap. No human can resist the chin on the knee, especially if you do the ol' melting puppy eyes bit. They will at least give you a pat on the head. Reward this with a little more chin pressure, and maybe a deep sigh. That should get them to rub your ears. Reinforce this behavior by leaning in, and maybe giving a little moan. Now turn your head enough that their hand slides to your neck or shoulder. Again, reinforce continued scratching by leaning in, and maybe looking up again with the puppy eyes. From here it should be a fairly simple matter to inch yourself forward until they are scratching that ever-itchy spot behind the hips. Heaven! And so simple. Just take it a step at a time, and reward every step of the way with the kind of expressions of adoration that humans crave.

Actually, the peoples might want to give some consideration to this technique. I notice that when they want a spouse or kid to do something that people get snarkier and snarkier trying to get the other person to comply. But maybe it would work more better to draw the person in the right direction with lovey-type stuff instead. I'm telling you, works for me every time.

Wags and kisses,

Thursday, May 27, 2010

Dominance, Cesar Milan, and the Corporate Control of America

While we're on hot-button issues for dog trainers, let's talk for a moment about dominance. There's a school of dog training, most famously represented at the moment by TV dog trainer Cesar Milan, which holds that the way to have a happy canine/human pack is to make sure that the human being is always in the dominant role. People who train from this perspective advocate things like making sure that the person always is the first to go through a door or the first to eat, and practice things like "alpha rolls" where the person pushes the dog into a "submissive" position belly-up on the floor. The kind of trainers who spend a lot of time talking about things like the timing and placement of reinforcers are not big on this particular theory, pointing out that a) social relations in wolf packs are a lot more subtle and complicated than these dominance trainers assume, b) dogs are not wolves and c) people are not dogs, so even if dogs do have dominance hierarchies, humans trying to create those the same way that dogs do might just confuse everybody. (Want to look tough? Try holding your tail stiff over your back. How's that working for you?)

OK, so I'm not a big fan of a lot of the dominance stuff, but it's also true that there are a lot of problems created in households where the dogs are running the show. If you can't sit on the couch when and where you want to sit, if you can't pick up your dog's food dish or take away something they want but can't have, there's a problem. Now that I think about it, there are a lot of households where the kids are running the show, which doesn't work any better. (If you can't set a bedtime, have homework and reasonable chores completed and get to school on time, to my mind, you've got issues.)

The solution to these problems, however, is not dominance, it's leadership. Leadership doesn't mean that I shove you around. Leadership means that I have access to the resources you want, and that you have to come through me to get them. Want the treats? Do the tricks. Want the TV? Get the homework done. Want access to the good things in life? Then follow the rules.

Here's where I go global (and maybe postal). For some reason, we seem to think it's OK for corporations, rather than the government, to be running the show. To my mind, that's like having the kids, or the dogs, in charge of the household. But now corporations have unlimited right to contribute to political campaigns. There's been deregulation of everything from banks to minerals management, and the rules that are in place are a) set by the corporations themselves and/or b) not enforced. And gee, it turns out that, like children and dogs, corporations really aren't qualified to be running the show. If the financial crisis was not enough to prove that point, the BP oil disaster ought to be driving the point home pretty good. If you want the right to drill for oil, you should have to conform to the safety rules. (Whether they can ensure that it's safe is a topic for another day.) If you want the government to loan you money, you should have safety limitations that prevent you from throwing people's money away on ludicrous gambling schemes.

Go ahead. Call me a socialist. But I would rather have a government elected by the people in the leadership role, rather than a bunch of corporations following their own interests without regard for public safety or welfare. Leadership through regulation and control of access to resources is all we've got, because I've yet to figure out how to alpha roll BP.

Monday, May 24, 2010

Bribes/Lures and the Great Oreo Question

OK, so nothing gets dogs trainers as up in arms as does the question of punishment. But you can get some pretty good (if somewhat less snarky) disagreement going over the question of just how you use rewards. For one group of people (clicker trainers) the reward always comes AFTER the behavior. It's, like, the definition of a reward. You do what I want, and I give you a reward. But how does the animal figure out what you want? Long story. We'll get there some other time. Suffice it to say that there is another group of people who find it useful to use the goody to show the animal what you want. For instance, if you have a puppy that you want to teach to sit, you can hold a treat just above its nose. The nose goes up and the bottom goes down. Puppy physics. Then the puppy gets the cookie.

Works just fine.'ve basically just bribed the dog. Rather than cheering after the fact, you've established the treat as the basis for getting the behavior. The puppy has lost the opportunity to use its own little brain to figure out the problem itself, and might not try it again until you pull out the treat to follow. Or, if you're on the other side of the argument, you've saved both yourself and the puppy the frustration of it wondering what the heck you want, by giving it a nice, clear, fun explanation. I told you there was disagreement.

But what, you ask, does this have to do with the wonderful world of people? And, yes, Oreos, which we were promised in the title? Well, it goes like this. Mattea (the kid) does a lot of dancing. Like ten classes a week. She and the girls she dances with compete in dance competitions and all. And the studio just added three new classes for kids her age with amazing teachers. To our surprise, however, Mattea didn't want to take the tap class with the ultra-cool, extremely hip and up-to-date tap teacher.

Mattea loves tap -- tap and hip hop are her favorite kinds of dance. However, unlike, seemingly, most of the girls she dances with, while Tea loves dancing, she has a limit. She likes to dance, but not if it takes away from her playing time. And this class was happening on a Tuesday evening, and she wanted her Tuesday evenings free. Now, from a parental perspective, we could see a variety of ways that Tea was going to get left in the dust if she blew off this class. But our policy has always been that dancing is a recreational activity, and recreational activities are things about which you get to choose (within the limits of the rules of the studio, which explains why she is taking ballet). So...we encouraged and we reasoned and we explained our concerns. And Mattea said no, she didn't want to.

And so Kelsey went to pick her up from hip hop, which immediately precedes this new tap class. But Kelsey brought with her a small package of Oreos. And she told Mattea that if, on the off chance, she were staying for the tap class, well then she'd probably need the cookies for extra energy. In other words, Kelsey bribed Mattea with cookies to take the class. Have I mentioned that Tea, like the dogs, is highly motivated by cookies? As in, pictures of Oreos on her computer background, just because it makes her happy to look at them?

Yeah, she took the class -- and the cookies. So, have we now ruined her intrinsic motivation to learn for the sake of learning? Or have we helped her over the hump to realize that the rewards for taking the class will exceed the costs? Will she take the class again? Will she demand cookies? Stay tuned...the class comes up again tomorrow.

Wednesday, May 19, 2010


It's coming up on the end of the school year -- always very exciting when you live with a kid and a high school teacher. But it's got me thinking about grades. I will avoid commenting on Mattea's grades in this public forum (but babe, if for some reason you should happen to read this -- since you DO your homework, is it that much more trouble to HAND IT IN?) But, you know, in a general, theoretical, totally non-parental, and certainly unconcerned way, what helps kids get good grades?

Karen Pryor, popularizer of behavior science and queen of clicker training, has an interesting post on her website about four schools that experimented with rewarding kids with money. As she summarizes the experiment:

New York set up a program to pay fourth-through seventh-grade children for their test grades during the school year. For great results you could get as much as $50. The money went right into a savings account.

Chicago also paid for test scores during the year. Good scores could earn up to $2,000 per year, half of which went into a savings account payable on graduation.

Washington, D.C. had a complicated system in which high school students were paid $100 every two weeks by getting perfect marks in five different areas, including attendance and good behavior.

Dallas kept it simple. Second-graders got $2 every time they read a book and passed a little computer test on it.

Care to guess which school had the best success? Right, Dallas. A test is an outcome, not a behavior. The way to change outcomes is to change behaviors. If you pay a kid $2000 at the end of the year for getting great grades or doing well on important standardized tests, all of the behaviors that would lead to success over the course of the year go unrewarded. Outcomes aren't really something that we can control -- behaviors are.

So if I were to promise Mattea $1000 for getting all A's on her report card at the end of the year (not gonna happen, babe!) she wouldn't really know what exactly she needed to do in order to achieve that goal. And the delay between the action and the reinforcement would be so long that she wouldn't really know what was being rewarded anyway. (And if I put the money into a saving account, it would hardly count as a reinforcement at all.) Far better for me to figure out what is standing in the way of her success and how to change that behavior. For instance, maybe I would need to give Tea a bag of M&Ms, and tell her that each time she handed in an assignment she could have an M&M. (If, of course, I could trust her not to gobble them all on the way to school. "Leave it" is a crucial command for both children and dogs. Mine are less reliable than they should be.)

How often have you talked to someone about their goal to lose 20 pounds, or to win a competition, or to learn to play the violin? All fine and good as far as it goes, but those are all outcomes, not behaviors. You can't get to the outcome unless you figure out what things you would actually need to DO to get to those goals. Much more effective to have a goal that you will eat fruit for dessert instead of ice cream, or practice for 20 minutes four times a week. Those are behaviors that you can control. And yes, that you can reward. (But you can't reward yourself for being good and eating fruit by having ice cream. That's cheating. It turns out that people very commonly reward themselves for exercising by eating more, or less healthfully, than they would otherwise. Wrong kind of reward.)

Personally, I would just like to say that this is the first year that I have actually fulfilled my New Year's resolution. That resolution was "More dancing!" (Always with the exclamation point at the end.) In addition to being a behavior, rather than an outcome, it has the advantage, as Taz would say, of being "more funner." So much easier to be successful at doing something that you wanted to do anyway.

What are your goals?

Monday, May 17, 2010

Punishment...and Crime.

If you ever want to see a big, snarly, hairy fight, you can toss a squirrel into the middle of a dog park, or you can bring up the concept of punishment in a group of dog trainers. Dog trainers run the gamut from "no kind of unpleasant thing shall happen to a dog on my watch, ever" to "if you don't shove the dog around he'll shove you around, so you better shove harder." I have no desire to become the squirrel at the dog park, so I'm just going to put out a few things that we do know about punishment and then ponder how they might apply to our penitential system.

So, what do we know about punishment? Well, for one thing, it's only effective if it happens right after the offending behavior. If you turn a hose on your dog as it's fence-fighting with the neighbor dog, it might (might!) think better of it next time. If you come home and find that your dog has chewed a corner out of your drywall or eaten a sofa cushion, your yelling or smacking will have no effect whatsoever, except to make your dog wonder why you have suddenly turned into a raving and dangerous lunatic.

We know that, unlike rewards, punishments need to happen every time the offending behavior occurs in order to be effective. If you sometimes get doused for barking and snarling at the neighbor dog and sometimes you just get to carry on, most likely you're going to keep carrying on.

We also know that punishment, especially if not skillfully applied, can create unforeseen associations and resentments/fears. If you hit a dog (or a child), you probably expect them to come out thinking "That was unpleasant, I should never try that again!" However, they may just end up with an impression of "Ow! That was nasty. Apparently I can't trust you not to hurt me." You can also end up with weird phobias, like the horse that Temple Grandin (Animals in Translation) talks about who is terrified of people in black hats after being hurt by someone wearing a black hat. You don't always create the same negative associations you planned on. about that crime and punishment? Say that you have someone who is using illegal drugs. Say that this person gets caught. How close in time is the punishment likely to be to the offending behavior? (Actually, I have no idea. I don't even watch Cops. But I know that people go for quite some time awaiting trial.) Is the punishment likely to happen every time the offending behavior occurs? Not a chance in hell. This one I'm sure on. Is putting this drug user in jail likely to create unintended negative consequences? It certainly wouldn't surprise me. I'm just kind of guessing that your average prison inmate doesn't spend most of their time reflecting on how prison is a logical consequence for their poor choices, and how they will never make such poor choices again. I could be wrong, but I'm kinda guessing that inmates are much more likely to be resentful of people in authority than otherwise. Not to mention that what small-time offenders are likely to learn from spending time with bigger-time offenders is hardly likely to be the kind of education you were hoping to sponsor.

Now, I'm not saying that we should abolish prisons. There are some folks who are simply too dangerous to have walking around, and they need to be contained. There's a difference between teaching and management. If your dog bites, put a muzzle on it before you go out in public. But the number of pathological serial killers/serial rapists, etc. has got to be vanishingly small compared to the number of people in prison. If prison doesn't work as punishment (punishment being defined as something that reduces a behavior) then aren't we spending a whole lot of money on something that simply doesn't work? (See here for one of many articles indicating that this is, in fact, the case.)

Nobody has asked my opinion on how to solve California's many-billion dollar budget deficit, but here is my modest proposal. Let everybody who isn't dangerous out of jail. Legalize drugs, since punishing people for taking them simply isn't keeping people from taking (or selling or producing) them. Have clear labels on all drugs indicating how dangerous they are, and levy high taxes. Use those taxes for drug treatment and prevention programs, and for public education programs about the hazards of drug use. These changes will, and the very least, save us many billions of dollars, and will probably result in lower rates of drug use.

Here's my question for politicians (and the people who vote for them) who are "tough on crime": does tough on crime mean "finding and implementing the most efficient possible ways to reduce crime" or does it mean "make people who commit crimes suffer"? Personally, I'd rather that we used the best information we have to get the best results, rather than assuming that our main goal should be to look as tough and snarly as possible. But hey, maybe I'm just the squirrel at the dog park....

Friday, May 14, 2010

Taz Talks -- More Funner

Hey Dudes and Dudettes! Mama says that she is thinking about a post on punishment, but she's not ready yet, so I can write this time. My topic is better. My topic for today is More Funner.

I will explain. I have a theme song. (Everyone should have a theme song!) Mine is "Tervs Just Wanna Have Fun." Actually, I think the lady originally sang it about Girls, not Tervs, but I'm not a girl, and I am a Terv (that's short for Belgian Tervuren), so it works better this way.

Tervs just wanna have fun. Dogs in general, really. People spend their time working and doing laundry and shopping for groceries and vacuuming and whatever, and when they have a bit of time left over, they figure that is their time for fun. Which they usually squander watching something stupid on the television -- although television can be an excellent way to keep people still for snuggling pets, which is, of course, crucial.

Dogs, on the other hand, are all about the fun (and napping). Yes, there are working dogs, with real, important jobs. But hel-LO! Check out the drug detection and search and rescue dogs -- big game of hide and seek, with games and prizes when you find what you're after. Herding sheep? Funnest game on the whole entire planet, no exceptions. K9 officers taking down bad guys? BAM! Human tug toy. Wicked awesome. Mostly dog jobs are a big ol' fun fest.

Now, I'm not saying that people jobs are necessarily the same way. Don't go all "following your bliss-y" and neglect to shop for dog food. But hey, why not make things More Funner? That's my point. With a little creativity, even the annoying stuff can maybe be More Funner.

For instance, Mama is always fighting with Mattea, the kid, over what stuff she should be eating. Mattea says that she never gets anything good to eat. Mama says that Mattea wants to eat nothing but junk food, and that bodies need protein and vegetables and not just empty carbs. (I'm not sure what empty carbs are. I am familiar with empty cartons, which are a whole lot of fun, but this is apparently not what she is talking about.) So a couple of days ago Mama made Chex Mix like stuff with the whole wheat cereal that isn't Chex because it's cheaper, and peanuts and almonds and some cheese crackers and spices and parmesan cheese and all of a sudden it's Party in a Bag for snack. Apparently it still has some empty carbs in it, but not all sugar and chemicals and stuff, and me and Coretta got to taste it and boy howdy is that stuff crunchy and delicious. Way More Funner than arguing about healthy snacks, but not so far off from healthy after all.

You want more kid examples? Unlike me, Mattea does not like to get up early in the morning. And she does not like people telling her to get up for school, but she also ignores the alarm clock. The other day Mama went in and sang to her the chorus from this old song:
Hal-an-tow, jolly lum-a-low
We were up
Long before the day-o
To welcome in the summer,
To welcome in the May-o.
For summer is a comin' in
And winter's gone away-o.

Mattea thought that was More Funner than just being told to get up for school. Now Mama sings the song this way:
Hal-an-tow, jolly buffalo
They were up
Long before the day-o
To welcome in the mustard,
To welcome in the May-o.
For buffalo are comin' in
So you better get out of the way-o.

Which is More Funner still. One more kid example. The other day Mattea was super mad about something, and told Mama that she wanted to hit her with something. So Mama got pillows out of the closet, and they both whammed on each other with pillows until Mattea was laughing and not mad. Personally, I found the sight of my people whacking each other a bit disturbing, but I guess the pillows turned out to be way More Funner than Tea yelling and Mama going on about Responsibilities and Self Control.

I think that if people want to make changes in their lives that they should find creative ways to make their lives More Funner. I will give you a hint. Stuff that you do with dogs is More Funner than stuff where dogs stay home and have a nap.

Wags and kisses,

Wednesday, May 12, 2010

In Which Lynn Wrests Back Control of her Blog to Talk More About Reinforcement

Here's the thing. I pay for the internet access. It's my blog. I don't mind sharing with Tazzie, who is, in all fairness, a more entertaining writer than I am, but I insist on at least getting my turn.

Now, where were we...ah yes, reinforcements (aka rewards). I believe before someone's long, pointy nose butted in, I was talking about how the properly applied use of rewards had resulted in a radically cleaner room for my daughter. It turns out that at this point in her 11-year-old life she is strongly motivated by money. Mostly because it will purchase the kinds of food that I refuse to buy. Some other time I will go on about just how much Mattea and Taz are alike.

But not now, because what I was really trying to get down to was the question of what kinds of rewards get results. People like money. Dogs like food. We know that, and we tend to rely on it in getting both people and dogs to work. But those aren't the only things that people and dogs like. People like food. Dogs don't much like money, although Coretta once ate a paycheck when she was a puppy. (Try explaining that one to the treasurer!") The crucial thing about reinforcements is finding out what is rewarding to the particular being that you're trying to motivate.

People quite commonly train dogs by rewarding them with games, such as tugging on a tug toy. And, of course, petting and praise. But you can also reward a dog by giving it permission to sniff something interesting, letting it chase you, scratching that special spot, throwing a ball or doing whatever goofy thing your dog finds entertaining. Sue Ailsby, a well-known trainer, says that giant schnauzers think it's hysterical when you try to grab their front paws. I wouldn't know -- I've never had giant schnauzers. Coretta hates having her paws touched, and Taz would prefer that you scratch the tops of his feet. A reinforcer is only a reinforcer if the one on the receiving end likes it. For some people, the biggest reward you can give them would be public recognition. Other people would rather have their teeth drilled than stand in front of a room full of people applauding.

So given that dog trainers have figured out such a wide range of ways to reward dogs, why do we not get that creative with motivating people? Sure, people work for money -- we need money, and we like to get more of it. But there are other things we like as well. It seems to me that if I were managing a group of people in an office, that it would serve me very well to figure out what everyone liked. "Wow, Jordan, thanks for putting in all the extra time getting this report ready. It looks great -- why don't you take the rest of the afternoon off?" "Jesse, that's the best idea I've heard all meeting! (Tosses piece of chocolate to Jesse)" "Pat, thanks for listening to my long diatribe about upper management. I know you'll keep what I said confidential. Hey, I was wondering if you had any new pictures of your grandkids to show me."

The great thing about using a variety of rewards with kids is that the proper use of timing turns things that they were going to get anyway into reinforcements for behaviors that you wanted. "Just as soon as your homework is done you can go over to your friend's house." "Once you've finished your vegetables you can have dessert." "If you get the dishes put away by 9:00 you'll be able to watch Glee."

There is, in fact, no reason why you can't use rewards to modify your own behavior. I'm sure I'm not the only one who tells myself "Two more paragraphs and you can play a game of solitaire." "Answer this tricky email, and then you can look at Facebook." I once knew a guy in seminary who would reward himself for completing a paper by giving himself permission to go use the bathroom. I'm all in favor of the judicious use of life rewards, but that's just sick.

What do you find rewarding?

Tuesday, May 11, 2010

Taz Talks -- Teaching People to Come When Called

Mama says she wants her blog back, but my friend Calli specifically asked that I write more about training people. Calli is a very smart dog (she's a Belgian Shepherd like me, but not fluffy), and I think we should respect her wishes. After all, if Mama is gonna take requests from her sister, why shouldn't I get to respond to my friends?

So here it is. Taz on training people: The Recall

Every dog wants their person to come when called, but people have a really annoying habit of ignoring you because they're "busy" at the computer or watching television or helping with homework or reading a book or whatever boring thing it is that people do when they're not paying attention to dogs. However, teaching your person to come when called is really very simple: Make a lot of noise.

Honestly, that's all there is to it. People cannot resist responding to big, agitated noise. Run to the door barking like your head is going to explode. Keep barking. They'll come check it out, guaranteed. And no matter how many times they say "You silly animal, there's nothing out there," or "What, all that noise about a squirrel?" they will still come. People just have a thing about big, agitated noise.

Don't believe me? Just turn on the talking radio or that Snooze Channel on TV. There's people who have shows that tons and tons of people listen to because those guys are barking super loud and agitated, and so people figure it's important. And they keep barking, so people figure what they say has to be true. It's like this -- Glen Bark: "OMG! OMG! OMG! There's a cat on the lawn! There's a cat on the lawn and it's gonna eat your grandma. The cat on the lawn is dangerous and scary and has big teeth and claws and it's gonna eat your grandma! It's a black cat, which is the scariest, most dangerous kind, and it might be after all a panther arrived here from Florida and it's definitely gonna eat your grandma!"

Oh wait, maybe that was me this morning. But Mama came when I called her like that, and millions of people around the world can be counted on to show up and pay attention if you just make enough noise and get all bent out of shape. It doesn't matter whether anything's actually there or not. Although there was a cat. I think. Maybe that was yesterday.

Wags and kisses,

Monday, May 10, 2010

Taz Talks

A-hem! It turns out that there has been a blog going on in my house for some little while -- a blog about training people, no less -- and I have not been invited to participate. Luckily, I'm not the kind of guy who stands on ceremony, and the Cheese Mom, who usually writes this blog, is otherwise engaged. (I am a lucky dog. I have both a Cheese Mom who is good at games with fabulous prizes and a Running Mom. Every dog should have this. Also an extra running partner called Peggy to take them on days when the Running Mom doesn't run. But I digress.)

What I was trying to say is that apparently this is a blog about training people, which is a subject that I know lots about. All dogs know plenty about training people -- starting tens of thousands of years ago when we started training you to feed us. I won't try to explain to you all of what I know about training people, as I expect I'll be able to get back to the computer at points future. For now, I thought I'd just pass on a few tips about a subject that people seem to be very interested in: food and exercise. Dogs are also very interested in food and exercise, but smarter, as indicated by the fact that you are the ones who have to go hunt down the food, while we get to eat it. Here are a few things I know that you should to:

  • Anything I can see, I will eat. Can you tell me that you are any different? If you don't want it to get eaten, hide it.
  • There are very few places you can hide food that I won't find them. You have opposable thumbs. Hiding it probably won't help. If you don't want to eat it, don't bring it in the house. (While there is plenty of stuff I can't get to, the cat can reach practically anything. And I make him share.)
  • We all need to chew on stuff. Dogs know this. People pretend otherwise. But seriously, don't you want to munch on stuff even when you're not hungry? Smart people give us dogs chew toys and bully sticks and kongs and stuff so that we don't eat the furniture. Why are there not better chew toys for people? Mama chews on gum. Coretta, my dog-sister, tried a pack and said it wasn't bad.
  • Exercise should be entertaining. For instance, running is best if you are chasing something, or something is chasing you. If you don't find chasing entertaining then find something else that's fun. People in my house like dancing. Also dogs in my house like dancing -- I'll put up video some time.
  • I get treats when I do something clever. How come people can just open up boxes and bags of treats and chow down without doing nothing for it? No fair. You should do something clever or useful before you have a treat. That way you will not only eat fewer treats, you will also be more clever and useful.
Uh-oh. The Cheese Mom is coming back for the computer. Gotta run. But hey, if anyone else has hot tips on food and exercise to share with people and humans, post a comment and let me know!

Wags and kisses,

Friday, May 7, 2010

Rewards: The Clean Room

Since I've only gotten one question in the brief life of the blog, it seems only fair not to let too long go by before I answer. My sister Lauren, quite rightly, wants to hear the story of Mattea (my daughter) and how I got her to keep her room clean. Which is a good story about reinforcement.

Let's start there -- what's reinforcement? Basically, it's a technical term for a reward. Even more technically, but usefully, a reinforcement is something that causes a behavior to happen more often. You smile at me. I smile back, so you're more likely to smile at me again in the future. When you're training animals, reinforcements are often food. You do a trick, you get a treat. But reinforcements can be anything that the animal/person who is learning wants. More about types of reinforcement some other time.

So here's the thing about how rewards work. They're only effective is they're associated with the behavior that you're trying to encourage. Obviously, if you hand me a cookie at 6pm, that won't necessarily connect in my mind with the fact that I graciously made my bed this morning. In order to communicate with animals about what exactly is getting rewarded, trainers often use a little noisemaker called a clicker to "mark" behavior. You hear the click, you know that whatever you were doing when you heard the noise is what is getting the reward. OK, but we're people. You could just tell me that the cookie is a reward for making my bed. That works, right? Maybe.

I really suspect that for people, as well as for (other) animals, the closer the reward is to the action, the more likely it is to have an effect. Here's where we get to the story of Mattea and the Clean Room.

Let me preface this by saying that room cleaning has always been a huge issue for us. Tea's room has always been a disastrous mess, and telling her to clean it up induced total melt-down screaming fits. "I can't! It'll take all day! No!!!!" Ick. Then we'd get mad, and she'd get madder and either she'd convince us to "help," which meant us doing all the work while she got distracted by each item she picked up or she'd pick up a few things and then be mad when we said she wasn't done. So we said, "OK, you can have a messy room, but you have to clean it up each Sunday." See above for what happened each Sunday. So I said to myself "You have studied all this behavior stuff. You know about using rewards. Reward her for cleaning." So I told her that if she kept her room completely clean, I would triple her weekly allowance (from $3 to $9). That seemed like a pretty good deal, since weekly room cleaning was already expected as part of the $3 allowance. The room was really clean for about a day and a half, but pretty quickly stuff piled up, and so no extra $6, and no change in behavior.

Here are the two really important things that I forgot. 1) Learning happens a little bit at a time. You cannot teach large changes, only a series of small changes. Much more on this another day. 2) Rewards need to be connected as closely as possible to the event they're rewarding. With an animal, that means within a second or two. With people you get a little longer, but a week was clearly not going to cut it.

So...I helped Mattea get her room well and truly clean. (Yes, she actually did work on it this time.) Once it was as clean as I felt it should be, I said that I would check every evening at bedtime. If the room was still this clean, she would get a dollar. Each evening was a new try. If you miss the clean goal one day, you can always fix it by the next day. (But you never make up the dollar you miss -- one clean day, one dollar, pure and simple.) It's been a couple of weeks now -- not exactly a lifetime habit at this point. But her room is strikingly clean, and she's gotten a dollar for every night. Even the night that she got mad because she couldn't find something and started throwing things around. She picked everything up before she went to bed.

Here's the interesting part. I don't always have a supply of dollar bills on me. So I said that I would write down the date of each evening that her room was clean, and that I would actually give her the money at the end of the week. So the actual timing of the reward is not particularly different than the version that didn't work. You could interpret this two ways. One possibility is that, as with the clicker, if you use a marker (my writing down the date) as a promissory note, and that marker happens promptly, you then have considerably more time to deliver the actual reward. Or, on the other hand, you could figure that the difference wasn't really the closer connection between the action (clean the room) and the reward ($$), but rather that the reward was based on a smaller behavior. "Keep your room clean for a week" wasn't a behavior that Mattea could master right off the bat. "Pick everything up before you go to bed" was something she could manage. I expect that success was based on a combination of the two (and probably a whole bunch of other stuff which I can neither identify nor control).

Better still, "pick everything up before bed" seems to have transformed into "try to put things away so that it's easy to have everything clean by bedtime." Here's the really great part. I was forever harassing her to hang up her towel after she took a shower in the morning. The only thing worse than damp towels lying about on the floor is me getting out of the shower to discover that now both of the towels are lying on her bedroom floor. Multiple ick. But if you leave a towel on the floor then your room is not clean, and you don't get a dollar. So she's been hanging up her towel.

Training new behavior: $7/week. Being able to leave her bedroom door open and having a dry towel when I need it: Priceless.

Wednesday, May 5, 2010

Name the Task Pt. 2: The Tea Party

I'm not quite ready to move on from the subject of identifying the problem before you try to arrive at a solution. For one thing, I didn't mention in my last post the crucial bit that you are only likely to come up with a solution if you genuinely want to solve a problem. My dog jumps up on me. People who teach dog training classes aren't supposed to let their dogs jump on people. But hey, he just kind of gently rises up and puts his paws on my chest for a scritch and a snuggle, and I think it's kind of sweet. I haven't trained him not to do it because I get something out of this particular "problem."

I'm wondering if this might be the case with the Tea Party movement. Clearly this is a group of people who are upset, and demanding change. But it's very difficult to tell what change it is that they're demanding. The official website gives as their mission/motto: "A community committed to standing together, shoulder to shoulder, to protect our country and the Constitution upon which we were founded!" OK, two missions: 1) protect country and 2) protect Constitution. So in order to accomplish the missions we would clearly need to determine who is threatening our country (Iraq? Iran? Al Qaeda? Illegal immigrants? Liberals?) and what is threatening our Constitution (Gun limits? Don't Ask Don't Tell? Free speech for corporations? Health insurance mandates?).

If their mission is to protect the Constitution, does that mean that they will stand up to protect same-sex marriage if the court (as in California) declares it constitutional? If a court should find that Arizona's anti-illegal immigration law is an unconstitutional violation of civil rights, will they stand up to protect the Constitution and the right of people to not be racially profiled? Something tells me this is not at the top of the agenda.

In reality, what the group seems to have greatest unanimity about is that taxes are too high, which seems only tangentially related to the group's stated mission. And even then, there doesn't seem to be any attempt to frame a solution as far as which taxes are too high, and what we might do without in order to lower taxes.

Which leads me to wonder, is this really a group in search of a solution, or a variety of solutions for a variety of problems? Or is the mission of the group really to protest and be outraged? In that case solving "problems" would actually run counter to their mission of being outraged.

Of course, the Tea Party is just one example of a problem that doesn't seem to be in search of a solution. I suspect that most of us have problems that we really aren't looking to solve. Some people complain all the time about being overworked, but actually love the feeling of importance and being needed that their overwork creates. Many, perhaps most, people bitch about their weight without having the slightest intention of permanently changing their habits of eating or exercise.

Part one of Name the Problem: If you don't know what you want to change, you aren't very likely to change it. Part two of Name the Problem: If you aren't trying to work out a solution, maybe you aren't really convinced you have a problem. Time to have another look at part one. Or maybe what I should call part 1a: If you call the problem something different (Protect the Constitution!) than the problem you'd really like to solve (I don't have a job), you are unlikely to come up with a solution. In which case cynics like me might suggest that you aren't really trying to fix anything after all.

Tuesday, May 4, 2010

Name the Task: A Dog Trainer's Take on Immigration Reform

So last post I wrote about how you get what you reward, not necessarily what you think you are rewarding. This brings me to what I believe is the most important step in changing/fixing anything. You have to figure out what it is that you want to change. It's just really awfully hard to teach something if you don't know what it is that you're trying to teach.

Say I have a cat who hangs out on the kitchen counters while I'm trying to cook. (This would be Finn McCool the Warrior Cat, who likes to snag broccoli out of the steamer, and tries to climb into pans. He's cute, but a pain in the butt.) I say to myself (and anyone within a two block radius) "The damn cat is in always in my food!" There. I've named a problem. The damn cat is always in my food. Now I can figure out how to solve the problem. I might toss the cat the into the gaping maw of the dog every time he got in my way (punishment), lock the cat in another room while I'm cooking (management) or teach the cat to sit on a chair while I'm cooking (training incompatible behavior). Pretty simple.

But most things in life are a little more complicated. All right, a lot more complicated. Take for instance The Immigration Problem. This is something that a whole lot of people are up in arms about these days. Depending on your perspective, the state of Arizona is currently either a courageous leader in tackling this crushing problem, or a brutal regime bent on racial profiling and civil rights violations. But here's the thing. What exactly is The Immigration Problem? Well, I guess we're probably talking about illegal immigrants, not legal ones. But still, what is the actual problem we're trying to solve? Is the problem that illegal immigrants are taking jobs from US citizens? Is the problem that illegal immigrants are using social services without paying taxes? Is the problem that dangerous drug lords and other criminals are getting in?

All of those possibilities are things that can be examined -- information can be gathered as to whether these are actual problems, or whether they are only perceptions of problems that don't really exist. And if they are, in fact, real problems, then the different problems would have different solutions. For instance, if the problem is that illegal immigrants are taking jobs that citizens want, it might make sense to crack down on employers who hire people who are here illegally. That solution, however, won't help at all with the problem of people using social services without paying taxes. (If, in fact, that is a problem that really exists.) You have to gather data to figure out the real problem(s) before you can do anything about finding solutions. Maybe it turns out that illegal immigrants, for the most part, are taking jobs that US citizens don't even want, and the solution is not to punish employers, but to give them a way to legally hire the workers they need.

The reality is that the world is full of incredibly complex problems. But the only way to start to solve them is to figure out what the problems really are. There is no such thing as The Immigration Problem. At least there's no possible way to solve The Immigration Problem. The Immigration Problem could be anything from "I was hit by an uninsured, unlicensed and undocumented driver" to "I'm terrified that my wife is going to be deported." The solution to "I'm uncomfortable being around people who speak a language I don't understand" is not the same as the solution to "If you're going to live in this country you should pay taxes." You can't solve a problem without clearly identifying a) what the problem is and b) whether your perception of the problem is born out by the facts.

Maybe the damn cat isn't always in my food. Maybe he's only in my food around the time that he expects to be fed, and I could easily get the cat out of the pan by feeding him his dinner before I make dinner for people. Maybe it's actually useful to have the cat in the kitchen cleaning up bits of things that I drop (which is, in reality, the dogs' job), and the problem is not that the damn cat is eating my food, but that the cat is getting on the stove, and I need to teach the cat to stay off the stove. The more I know about the problem, the better the chance I have of coming up with a solution.

You get what you reward, not what you think you're rewarding. You can only solve actual and particular problems, not generalized perceptions of problems.

Monday, May 3, 2010

A Dog Trainer Explains the Financial Crisis

OK, so here goes...

As I mentioned last post, behavior scientists have pretty well figured out that you can encourage a particular behavior by rewarding it. Go figure. Dog sits, you give it a treat, and dog is likely to sit again. Big financial corporations, as far as I can tell, have figured out the same principle. Banker-type makes money for the company, and the company gives Banker-type extra money as a reward, to encourage making more money. All very sound. You can think of it as training Fat Cats, rather than dogs. (Oh, and by the way, these training principles work just as well with cats as with dogs. Don't believe me? Come over and I'll have my cat do tricks for you. Of course, they also apply to chimps, horses, llamas, dolphins, fish and people. That's the point.)

Sorry...where were we? Oh yes, Fat Cats getting money for making money. OK, so where's the problem? Goldman Sachs, say, rewards folks for making money by giving them money. Success is rewarded -- capitalism at its finest. But here's the thing. The law of behavior says that you get what you reward. What you reward, not your idea of what you're rewarding. Say you have a dog that barks in an annoying fashion. Every time that dog barks, you yell at it. Well, if the dog happens to enjoy the fact that now both of you are yelling together, then you have rewarded the dog for barking by joining in on its yapping. Then you get both more yapping and a sore throat.

But making money is a good thing, right? That's what financial institutions are for is making money. Where's the problem with that?'s the thing. Legitimate businesses provide goods and/or services. Investment banks are supposed to make money for their investors. Business that just make money for themselves, without providing any goods or services in exchange, are known as things like Ponzi schemes, counterfeiters and bank heists. Bernie Madoff for sure had one these kinds of businesses. Financial institutions were rewarding people not for making money for the customers, but for making money for the bank. So Goldman Sachs (for instance), sold financial instruments to customers, knowing that they were likely to end up worthless, and then bet in the markets that those financial instruments would tank. So the bank made money, the bankers made money and the investors (and the rest of us who have been affected by this economic mess) lost out. The bankers were rewarded for making money, so they made money -- by essentially lying and cheating. Mortgage brokers were rewarded for selling mortgages. Not for helping people to get appropriate mortgages that would make it possible for them to own homes over time. Brokers were rewarded for selling mortgages, and the bigger the mortgages the bigger the reward. Until it turned out that people couldn't pay the mortgages that they'd taken out, and the whole company (say, Countrywide) collapsed, along with the housing market. Ultimately, making money by making money is not sustainable.

What if, instead, businesses had clearly and publicly articulated missions (which I presume they do), and they figured out a way to reward people for advancing the overall mission of the business? Better yet, what if bonuses were tied to particular innovations that led to measurably better performance in particular facets of the business's mission? (More on this distinction later.)

Here is my assertion #1: You get what you reward, so you should be pretty damn careful about what exactly it is that you are rewarding.