Kelly S, Green G, Sidman M
New England Center for Children, Southborough, Massachusetts 01772, USA.
J Appl Behav Anal. 1998 Summer;31(2):237-43. doi: 10.1901/jaba.1998.31-237.
After preliminary computerized training on visual-visual identity matching, a 5-year-old boy with autism (Sam) was given visual-visual and auditory-visual matching-to-sample tests with new stimuli. He did well in matching dictated name samples to 20 pictures, 26 printed upper case letters, and 9 single-digit numbers. In matching the visual stimuli (pictures, letters, or numbers) to themselves, however, he did not perform well. We then increased the number of picture comparisons per trial from two to three. In tests after this three-comparison training, Sam correctly matched on 95% of the original 20-stimulus, four-comparison, identity-matching test trials. He went on to demonstrate accurate identity matching of the numbers, letters, and new pictures. In identity-matching tests on the table top, he performed poorly until the stimulus array was made to resemble the stimulus arrangement on the computer. These findings showed that seemingly small procedural changes can influence performance and demonstrated that successful auditory-visual matching does not guarantee proficiency in visual-visual identity matching.
在对视觉-视觉身份匹配进行初步的计算机化训练后，一名患有自闭症的5岁男孩（山姆）接受了使用新刺激物的视觉-视觉和听觉-视觉样本匹配测试。他在将听写的名字样本与20张图片、26个印刷大写字母和9个一位数数字进行匹配方面表现出色。然而，在将视觉刺激物（图片、字母或数字）与它们自身进行匹配时，他表现不佳。然后，我们将每次试验的图片比较数量从两个增加到三个。在这次三次比较训练后的测试中，山姆在原来的20刺激物、四次比较的身份匹配测试试验中，有95%的试验正确匹配。他接着展示了对数字、字母和新图片的准确身份匹配。在桌面身份匹配测试中，直到刺激阵列被调整得类似于计算机上的刺激排列，他的表现都很差。这些发现表明，看似微小的程序变化会影响表现，并证明了成功的听觉-视觉匹配并不能保证在视觉-视觉身份匹配方面的熟练程度。
