many people assume that punishment stops undesired behavior. is this always true? psychologists have learned that the effect of punishment depends greatly on its timing, consistency, and intensity.
punishment suppresses behavior best when it occurs as the response is being made, or immediately afterward (timing), and when it is given each time a response occurs (consistency), thus, a dog that has developed a habit of constantly barking can be effectively (and humanely) punished water is sprayed on its nose each time is barks. ten to fifteen such treatments are usually enough to greatly reduce barking. this would not be the case if punishment were applied occasionally or long after the barking stopped. if you discover that your dog dug up your flower bed while you were gone, it will do little good to punish him hours later. likewise the commonly heard childhood threat, "wait until your father comes home, then you'll be sorry" does more to make father an ogre then it does to effectively punish an undesirable response.
TANATiP
severe punishment can be extremely effective behavior in stopping behavior. if a child sticks a finger in a light socket and gets a shocks, that may be the last time the child ever tries it. more often, however, punishment only temporarily suppresses a response. if the response is still reinforced, punishment may be particularly ineffective. responses suppressed by mild punishment usually reappear later. if a child sneaks a snack from the refrigerator before dinner and is punished for it, the child may pass up snacks for a short time. but since snack sneaking was also rewarded by the sneaked snack, the child will probably try sneaky snacking again, sometime later.