PositiveReinforcementVsBribing

I have been wondering lately about the difference between Positive Reinforcement (which increases motiviation and creates enthusiasm) and Bribing (which is known to decrease motivation and create a sense of LearnedHelplessness). I am not sure I understand where to draw the line. Anyone have any ideas?

I would not be surpised if this issue was also tied up in the distinction between InchPebbles and MicroManagment which are also supperfically similar but have different motivational concequences.

KenEstes 2003.02.05

Having temporarily mislaid my copy of "Don't Shoot the Dog", I'll have to use a cutesy definition:

After a long time, if you must give the reward in order to get the behavior, then the reward is a bribe.

If you don't have to give the reward to get the behavior, then the reward was positive reinforcement.

:-)

--DaveLiebreich 2004.02.05

I think Dave's got it. A bribe comes before the behaviour. Positive reenforcement comes after. Ken's question still extends for me. Can positive reenforcement encourage future bribery? If yes, how do you prevent it?

MikeMelendez 2004.02.05

There's also a narrow effectiveness of motivation kind of issue. People perform better with intrinsically motivated tasks. Adding extrinsic motivation - rewards or bribes - to an intrinsically motivating situation actually decreases performance.

There was some wonderful work done on this, published for the general population in Why We Do What We Do by Deci and Ryan. That particular book, and the research it draws from is interesting because these guys are former behaviorists, especially Ryan. So their experimenal design, analysis of data, and so on is rigorous compared to most of the stuff in the "rewards are dysfunctional" literature.

- JimBullock, 2004.02.05

Jim said, "People perform better with intrinsically motivated tasks. Adding extrinsic motivation - rewards or bribes - to an intrinsically motivating situation actually decreases performance."

A case made in detail in the book Punished by Rewards by Alfie Kohn

EstherDerby 020504

Wow, everyone is familiar with the books that I was thinking of.

I am still not sure I understand the difference. The way Karen Pryor uses positive reinforcement, she claims that it makes the animals more inquisitive and eager to try things out. Why does this not go against the research of Deci and Ryan who show that once you start giving rewards people become demotivated to work on the task. My thought is that there might be some subtle differences between the way Karen trains her animals and the way the experiments "pay" their subjects. There is no doubt in my mind that a reward of a pat on the back can be reinforcing while money can sometimes be demotivating. I am wondering if the difference is in presentation and connection with the person rather then the size of the reward.

To be fair Karen also uses positive reinforcement to be demotivating. (this was my favorite part of the book). She had one dolphin who had a behavior she did not like (sink to the bottom of the tank and refuse to "worK") so she started to "pay" the animal for this behavior, she turned the unwanted behavior into work so that the animal refused to be obstinant unless it got paid!

Jim got right to the heart of the issue:

There's also a narrow effectiveness of motivation kind of issue. People perform better with intrinsically motivated tasks. Adding extrinsic motivation - rewards or bribes - to an intrinsically motivating situation actually decreases performance.

YES!

but is it not true that the positive reinforcement methods ala skinner are effective in teaching new behavior? How do you try to intrinsically motivate a dog?

KenEstes 2004.02.06

A thought: Animal trainers usually have control over most of the animal's environment. Your positive reinforcement of a coworker has to compete with the rewards system of the culture.

--DaveLiebreich 2004.02.06

One reason for your question, Ken, is definitional. The several sources are talking about slightly different things, from slightly different perspectives.

As for this: "How do you intrinsically motivate . . . " Dogs are intrinsically motivated - very much so. Dogs do all kinds of things for which there is no payoff other than whatever payoff occurs inside the dog, in doing just that activity, without regard to whether someone else will give them a payoff. People are intrinsically motivated as well.

- JimBullock 2004.02.06 (Why did I just do that?)

Dave, I think you underestimate the richness of the dog's environment, as well as how narrowly focused some humans can be. The dog may not be motivated by the same things that your co-workers are, but believe me, there's no end of things that dogs find rewarding.

And, of course, each being--dog or human--is an individual. So, when you're training by positive reinforcement (or bribing, for that matter), you'd better first "calibrate" the trainee to discover what rewards them. That's an essential part of dog training and all other forms of experiential training. In neither case are the trainees motivated by lectures. - JerryWeinberg 2004.04.06

Or put another way, animal trainers are lucky to have enough control over an animal's environment that the animal will pay any kind of attention at all. Sometimes a tiger gets discracted by someone's bad hair-do, and drags you out of harm's way, whether that's what you want or not. This is the current story for what happened on-stage when the magician "Roy" of "Sigfried and Roy" was hurt by one of the tigers in his act. I don't know about the hairdoo thing. I do believe that with the hold the big cat had, if it were trying to harm Roy, he wouldn't just be dead right now, there would be no head on his body.

-- JimBullock, 2004.01.06 (Watch out for those beehive hair-doos.)

Jerry,

Do you have any insight as to when positive reinforcement becomes demotivating?

KenEstes 2004.02.08

When it's poorly done--which could be

bad timing of the reinforcer, like too long after the event you want to reinforce (as in annual personnel reviews)
poor choice of reinforcer (one that isn't motivating for this trainee, like money offered in reward for a voluntary action)
reinforcing a behavior that looks like what you want, but is actually a different behavior

Others can add more examples. These are the ones I see most often, I think. - JerryWeinberg 2004.02.08

Updated: Monday, February 9, 2004