Diamond A
University of Pennsylvania, USA.
J Exp Child Psychol. 1995 Jun;59(3):419-56. doi: 10.1006/jecp.1995.1020.
Infants of only 5-6 months prefer to look at something new when given the choice of looking at a stimulus shown earlier or something new, even after a long delay (the visual paired comparison task). However, if infants must reach and displace a stimulus to retrieve a reward, even 18-month-olds respond randomly when given the choice of reaching to the stimulus shown earlier or to something new, even after a brief delay (the delayed nonmatching to sample task). To investigate this paradox we modified the delayed nonmatching to sample task to make it more similar to visual paired comparison. Each stimulus served as its own reward; no rewards were hidden under any stimuli. Infants were habituated to a sample object, a delay was imposed, and then the sample and a new object were presented. Infants could choose to look at (in visual paired comparison) or reach for (in delayed nonmatching to sample (stimulus = reward)) either object. One hundred twenty infants were tested: 60 (20 each at 4, 6, and 9 months) on visual paired comparison and 60 (20 each at 6, 9, and 12 months) on delayed nonmatching to sample (stimulus = reward). The same 10 pairs of stimuli were used on both tasks. Each subject was tested twice at all five delays (10, 15, 60, 180, and 600 s). At even the youngest age that reaching was tested (6 months), infants showed evidence of recognition memory on the reaching task at delays at least as long as those at which they demonstrated recognition memory on the looking task. Indeed, when subjects reached, not in order to obtain something else, but to obtain the stimulus itself, they succeeded on a recognition memory task even at delays 10 min long very early in life.
在视觉配对比较任务中,仅5至6个月大的婴儿在有机会选择观看先前展示过的刺激物或新的刺激物时,即使经过长时间延迟,也更倾向于观看新的刺激物。然而,在延迟非匹配样本任务中,如果婴儿必须伸手移动一个刺激物以获取奖励,即使是18个月大的婴儿,在有机会选择伸手去拿先前展示过的刺激物或新的刺激物时,即使经过短暂延迟,他们的反应也是随机的。为了研究这一矛盾现象,我们对延迟非匹配样本任务进行了修改,使其更类似于视觉配对比较任务。每个刺激物本身就是奖励;没有奖励藏在任何刺激物下面。婴儿先对一个样本物体形成习惯化,施加延迟,然后呈现样本和一个新物体。婴儿可以选择观看(在视觉配对比较任务中)或伸手去拿(在延迟非匹配样本任务(刺激物=奖励)中)任一物体。对120名婴儿进行了测试:60名(4个月、6个月和9个月各20名)参与视觉配对比较任务,60名(6个月、9个月和12个月各20名)参与延迟非匹配样本任务(刺激物=奖励)。两项任务使用相同的10对刺激物。每个受试者在所有五个延迟时间(10秒、15秒、60秒、180秒和600秒)下都接受了两次测试。即使在测试伸手动作的最小年龄(6个月)时。婴儿在伸手任务中也表现出了识别记忆的证据,延迟时间至少与他们在观看任务中表现出识别记忆的延迟时间一样长。事实上,当受试者伸手去拿不是为了获得其他东西,而是为了获得刺激物本身时,他们在生命早期就成功完成了一项延迟长达10分钟的识别记忆任务。