“You can catch more flies with honey than with vinegar.”

Everybody’s Grandma

 

“Positive reinforcement is defined by what the other person does, not by what you do.”

Aubrey Daniels, author of Other People’s Habits

 

When my children were at the potty training age, I looked high and low for a good guide to potty training.  There weren’t many to choose from.  One book suggested giving a child an M&M every time he sat on his potty seat and claimed that pretty soon he would begin to relieve himself there.  Right.  I doled out quite a few M&Ms in the bathroom, but my first child did not seem to have any clue that there was a relationship between the candy and his bodily functions, or else he didn’t care.  (And since he really preferred fruit to chocolate, I ate way too many M&Ms in the process!)  The worst thing was that pretty soon he began to resist going into the bathroom at all.  Clearly the book overlooked teaching me some fundamental details about positive reinforcement.    

                “Oh, I tried positive reinforcement.  It didn’t work.”  I have come into contact with several people who claimed to have given up on clicker training with their dogs or on using positive reinforcement in changing their children’s behavior because they tried it and nothing happened.  But did they really use positive reinforcement?  Or were they just doling out M&Ms? 

            The definition of positive reinforcement states that a behavior is followed by a consequence called a reinforcer and results in a future increase in the rate of that behavior.  The consequence, of course, must come after the behavior happens.  And the consequence must cause the behavior to happen more often in the future.  Otherwise it’s not positive reinforcement in the first place. 

            A reinforcer can be anything that follows a behavior and makes the behavior’s rate increase.  It’s not just food.  It’s not just toys.  And especially it’s not just praise.  It could be any of those, and just about anything else you can think of.  Every individual has a unique collection of experiences that will reinforce certain behaviors for that individual.  One reinforcer may not work in every situation or at different times in the learner’s life.  And one individual’s reinforcer may seem just weird as can be to someone else.  It has to be something that really works to increase the rate of behavior for that person in that situation to be called a reinforcer.   

            The consequence must come after the behavior happens. In every day lingo, the term ‘consequences’ is sometimes loaded.  When we hear someone say, “There will be consequences!” we have learned to expect something unpleasant to happen.  But consequences are really just events that happen after and in relation to some behavior.  So the behavior and the consequence have to be related in time.   

            The behavior must happen more often in the future.  More importantly, the consequences must result in an increase in the behavior.  If the behavior doesn’t increase, no positive reinforcement took place.  You can’t accurately say it didn’t work if it didn’t even happen to begin with. 

            But why wouldn’t giving a child an M&M increase his doing his business on the potty, especially if he likes and will eat M&Ms?  Why wouldn’t giving a dog a Milkbone increase his sitting on cue?  There are several things to consider.  In his directed learning textbook for college students, Principles of Everyday Behavior Analysis**, Keith Miller describes four principles of effective reinforcers.  They are Deprivation, Immediacy, Size and Contingency.  To make it easier to remember, their first letters spell the word DISC.

            Deprivation.  In order to be effective, your learner must be deprived of the item you plan to use as a reinforcer.  When we talk about deprivation, it doesn’t mean that he has to be starved for 2 days before you can train him.  It just has to mean he hasn’t had so much of the thing you’re offering him that he’s not tired of it, too full or just not interested in it any more.  Your dog may love Scooby Snacks, but if he’s just eaten the whole pot roast you left on the kitchen counter, chances are he’s not going to be that hungry.  He has to be at least a little bit deprived of the thing (or type of thing) you’re offering him.  If your dog has just eaten a lot of food, you’ll have to find some other kind of reinforcer to use that he hasn’t just had too much of.  Attention, a favorite toy that he doesn’t usually get to play with, going on a walk or riding in the car could work.  If your learner is not deprived of one reinforcer think of another thing he likes and see if that works.

            Immediacy.  The sooner the reinforcer is delivered after the behavior, the better chance there is of it being effective.  If you deliver a reinforcer too late, you might accidentally reinforce what the learner was up to when you finally got the prize delivered.  For example, if a student, Little Kim finally succeeds in getting a math problem right and her teacher gives her a sticker 5 minutes later when she is talking to Little Josie, what behavior is most likely to be reinforced?  Yep.  Talking with Josie.  Not the behavior the teacher wanted to see increased. 
            As adults we are used to waiting for our reinforcers.  We work for a week or two… or even a month… before we see a paycheck for our efforts.  As a freelance writer I've sometimes waited 7 or 8 months, or even a year to be paid for writing I've done!  The paycheck works to keep us going to work.  But with young learners and animals time is of the essence.  The truth is that with adult learners the very same thing is usually true, even though we’ve learned to wait for reinforcement in some situations when the reinforcer is really valuable.  We may persist at working for pay that won’t show up for several weeks, but what happens to our day-to-day work quality if that’s the only reinforcer we get?  If you want the reinforcers you deliver to be effective, get them to your workers and learners as fast as possible.

            Size.  Well, you have to give Dr. Miller credit for finding a word that works in the DISC acronym, but size isn’t all we’re talking about here.  Size is only part of it.

            Size refers to how worthwhile the thing we want to use as a reinforcer is for the learner.  How large or how many pieces of a reinforcer there are is part of it.  Most learners will choose the larger reinforcer.  If your horse won’t work for a 1/2” slice of carrot, try a 1” slice.**** 

            But size isn’t the only thing that makes a thing worthwhile.  If you’re using an activity as a potential reinforcer, it should be something the learner enjoys doing.  If it’s food, it should be food the learner likes a lot.  (In my potty training example above, had my overall training skills been better, I probably would have had better luck training with bits of fruit rather than M&Ms with my son.) 

            But liking it ain't enough.  Even if your learner really likes what you’re trying out as a reinforcer, it has to be something he is willing to work for.  If he’s not willing to work for it, it’s not a reinforcer.  Sometimes a learner will appear to be quite fond of something, but still not work for it.  In that case it isn’t a reinforcer.  Try something else! 

            Contingency.  This is an important one!  When we’re talking about behavior, contingency refers to a relationship between behavior and the consequence.  What Miller says is that the consequence has to happen if and only if the behavior occurs.  If a teenager does the dishes and his mother says, “Thank you!  Now you can watch TV!” the television viewing won’t be very effective as a reinforcer if the kid has free access to the TV any time he chooses to watch it.  He can always watch TV, so what’s the big deal about doing the dishes?  But if there is a family rule that TV viewing is doled out in relation to performance, and if the child enjoys watching TV (remember the size/worthwhile principle!) then it is much more likely to act as a reinforcer. 

            The bottom line is that positive reinforcement always works.  You just have to be sure that what you’re delivering is really positive reinforcement.     

 

Copyright, 2004  Kellie Snider

December 12, 2004

 

** Miller, L. Keith, Principles of Everyday Behavior Analysis, Wadsworth Publishing; 3rd edition (May 31, 1996)

*** Not all, however!  Sometimes a learner will consistently choose the smaller portion of a reinforcer over the larger portion.  This is not common, though. 

**** Watch out with reinforcer size, however!  If your reinforcers are too large your learner will become satiated and unless you have other effective reinforcers at hand your training session will come to an end.