Huguenin N H, Touchette P E
J Exp Anal Behav. 1980 Jan;33(1):77-86. doi: 10.1901/jeab.1980.33-77.
Eight severely retarded young men learned color and line-tilt discrimination. After 95% accuracy was achieved for both dimensions, they were combined to form "conflict-compound" stimuli in which prior reinforcement history was reversed for one element of the compound and unchanged for the other. When responding to the compound was 95% accurate, control exerted by each element was measured. The unchanged element consistently exercised control in agreement with the reinforcement contingencies associated with the compound, regardless of whether it was color or line orientation. The reversed element, which had a conflicting prior history or reinforcement, most often exerted control associated wtih original training, or no control, suggesting that it had been "ignored" during the compound. Conflict compounds produced selective attention. When elements were combined to form "compatible-compound" stimuli, both exercised control in agreement with the compound in post-tests. Selective attention was not produced by compatible compounds.
八名重度智障青年学会了颜色和线条倾斜度辨别。在两个维度的准确率都达到95%之后,将它们组合形成“冲突复合”刺激,其中复合刺激的一个元素的先前强化历史被颠倒,另一个元素保持不变。当对复合刺激的反应准确率达到95%时,测量每个元素施加的控制。无论不变的元素是颜色还是线条方向,它始终按照与复合刺激相关的强化条件施加控制。具有冲突的先前历史或强化的颠倒元素,最常施加与原始训练相关的控制,或者不施加控制,这表明它在复合刺激期间被“忽略”了。冲突复合刺激产生了选择性注意。当元素组合形成“相容复合”刺激时,在测试后两者都按照与复合刺激一致的方式施加控制。相容复合刺激不会产生选择性注意。