Explain errorless discrimination Errorless Discrimination Learning…

PrivateWater38693Explain errorless discrimination Errorless Discrimination Learning…Explain errorless discriminationErrorless Discrimination Learning and Easy-to-Hard TransferChapter 4 lays out the principles of error-correction learning, where learning in the Rescorla-Wagner model is driven by prediction errors—that is, by the mistakes in prediction that an animal makes when it does or does not anticipate an unconditioned stimulus (US). Chapter 4 also reveals that errors need not be the key to all forms of learning. In fact, as we will soon see, it is possible to develop training regimens that all but avoid errors—and, in doing so, to enable animals (including people) to learn very difficult discriminations with little or none of the frustration and emotional angst that come from making mistakes (emotions you might have experienced yourself on the last test or quiz you took).Herbert Terrace developed a training procedure he called errorless discrimination learning. Training begins with a discrimination task that is readily learned and then transitions to a similar but different discrimination task that is harder to learn so as to eliminate, or drastically reduce, the errors made during the harder task.errorless discrimination learningA training procedure in which a difficult discrimination is learned by starting with an easy version of the task and proceeding to incrementally harder versions as the easier ones are mastered.In one study of errorless discrimination learning, Terrace trained pigeons to discriminate between a red disk (S+) and a green disk (S-) and peck only the red disks (Terrace, 1963). Normally, this would take many trials and produce many errors, much as in the classical and operant conditioning curves shown in Chapters 4 and 5. Terrace got the pigeons to learn this discrimination quickly, and with almost no errors, by giving them a hint—like a set of “training wheels” for learning—to keep them from making the wrong choices. Early in training, he made the red disks very bright and bold (to naturally attract the pigeons’ attention) and kept them available for 3 full minutes. In contrast, the green disks were presented as dark or dimly lit and were available for only a few seconds. Since pigeons generally do not peck at dark objects, especially if they have limited time to do so, the pigeons rarely pecked the wrong green S- keys. Gradually, over many trials, Terrace increased the brightness and duration of the green S- keys until they were as bright and lasted as long as the red S+ keys. Terrace’s pigeons were now able to discriminate the two colors without the extra hint and guidance of the modified duration or brightness of the keys.Terrace’s errorless learning procedure has had many applications in education but has seen its greatest success in training people with learning disabilities. By means of a procedure similar to the one in the Terrace study, children with Down syndrome were taught to identify basic shapes such as ovals and rectangles (Duffy & Wishart, 1987). This procedure involved flash cards that initially contained only the correct shape for the name given (so no errors were possible). Then the children were shown cards containing two choices, one being the correct shape in a large format and the other being the incorrect shape set very small. The very small size of the incorrect choice guided the child to ignore it, akin to the effect of dimming the lights in the Terrace study; thus, the child was naturally inclined to choose the larger option. Over time, the sizes of the two alternative choices were made increasingly similar until they were the same (much as Terrace gradually made the lights for the pigeons the same brightness and duration), and the children with Down syndrome ultimately learned to identify shapes correctly—avoiding the errors and frustration that would have resulted had they been given the final test at the very start.Errorless discrimination learning does have significant drawbacks, however. While it produces rapid and strong (and painless) learning of the discriminations being trained, many later studies showed that this learning is very rigid and inflexible, adhering to the details of the original training procedure rather than generalizing to new situations and stimuli not trained in this fashion (Clare, Wilson, Carter, Roth, & Hodges, 2002; Jones & Eayrs, 1992). Nevertheless, in the study described above, the errorless discrimination learning, by preventing errors, also prevented the frustration, trauma, and self-doubt the children with learning disabilities might have suffered from making many errors at the start. Like a parent who doesn’t want to see a child fall and hurt himself trying to ride a two-wheeler before he is ready, the errorless learning procedure offers a gradual form of learning that begins with an easy task and then incrementally moves to the harder version when the person is ready and able.Social SciencePsychology