spiritdogs
Posted : 10/31/2006 6:37:30 PM
ORIGINAL: mrv
If positive reinforcement did not "work" then positive reinforcement being used by the person involved was not effective. If the behavior is being maintained at a consistent level, then something in the environment is positively reinforcing the undesired behavior. So you need to look at the whole situation from prior to the behavior to after the behavior occurred. A video can be very helpful.
Another reason it will be stated that "positive" reinforcement did not work dealt with issues of fading the supports that resulted in behaviors that were targeted. If a kid can spell words on a spelling test only when given in a certain order, can the kid really spell the words. There needs to be a specific plan to fade cues or supports, lengthen the time between behavior and reinforcement and fade the reinforcement to a much leaner schedule.
The techinque was not properly applied; it will work if the situation is examined and a more specific plan with greater attention to the environmental issues and the shaping of behavior and fading of supports needs to occur.
I totally agree. Positive training takes time, patience, and proper technique. It rarely fails, properly applied.
Unfortunately, if a dog is still jumping on humans, probably what has happened is that the dog has been accidentally reinforced for doing so. Reinforcement comes in many forms, and even a negative can be perceived by the dog as reinforcement. As can eye contact, shoving him away, making a noise at him, etc. I have found that ignoring is really the best thing you can do, but you must do it consistently, with no voice, no eye contact, no nuttin' until the behavior extinguishes. The thing that gets most people in trouble is the extinction burst - the dog's behavior gets worse (after all, whatever he was doing
used to work, so he tries harder at it), so they assume positive training didn't work. But, had they continued, and waited it out, still ignoring, the next thing that follows the extinction burst is - well, extinction.
One can do things to try to create a situation that makes the dog successful. Example, the dog that pulls like crazy. If you acclimate the dog to a Gentle Leader, or Halti, you can control his head, which means he no longer pulls you off your pins. Then, you set about your training regimen. Walk two steps C/T (click/treat, or click/reinforce), walk another two steps C/T. After the dog is consistently doing that, you can add a cue word or phrase, like "walk with me" or "let's go", or "heel".
Another mistake people make is that they put the dog into situations it cannot yet handle. If you raise the distraction level, you must temporarily lower your expectations. And, if you are trying to desensitize a leash reactive dog, for example, that barks at dogs that approach within a certain distance, you begin your work just at the point where the dog appears, but is not yet at the point where it elicits reactivity from your dog,
no matter how far away that is. But, often, people say positive didn't work because their dog wouldn't take a treat. Exactly - the dog is telling you he's too close to the scary stimulus - you simply need to increase the distance until the dog is comfortable taking the treat. Then you know where to start!