自闭症儿童利用社会性注意线索习得词语的眼动研究

本研究采用眼动技术考察18名自闭症儿童利用社会性注意线索习得词语的能力。结果表明：1）自闭症儿童具有利用他人视线习得词语的能力；2）不同视线对自闭症儿童的词语习得产生不同的影响；3）自闭症儿童视线追随行为的潜在机制与普通儿童存在差异，自闭症儿童的视线追随行为可能是意识性的诱发行为而普通儿童则是反射性的自发行为。

Word Learning with Social Attention Cues in Children with Autism: Evidence from eye movements

Many literatures about word learning argue for the importance of joint attention, while previous researches indicate that children with autism disorders (AD), unlike their typically developing peers, do not automatically orient their attention to social attention cues. Then, do deficits in social cognitive skills impede children with AD to capture social attention cues to learn words? This research investigated the ability of learning words via social attention cues in verbally-able children with AD by eye-tracking technology. We recorded the eye-movements of 18 verbally-able children with AD (M ages=134.83 months) and their typical peers matched on verbally and non-verbally intelligence (M ages=71.61 months), while completing word learning tasks in three conditions: in the No Gaze condition, the speaker in the video straightly gazed at and faced the camera; in the Subtle Gaze condition, she gazed at the target object and straightly faced the camera; while in the Salient Gaze condition, the speaker not only gazed at, but also faced the target object. Behavioral data showed that children with AD were able to take advantage of gaze cues to detect the referents of novel words no matter whether gaze cues were accompanied with head direction or not, which indicated that AD children had the same proficiency of word learning skill via gaze cues as their typical peers did. Eye movement data results showed that: 1) the frequency of gaze following (the times of saccades from face to target objects) in AD children were lower than TD children; 2) the proportion of duration on eye region (fixation to eye region/ fixation to face region) in AD children were higher than TD children; 3) there were no differences between two groups in proportion of fixation to the target objects (fixation to target objects/ fixation to target and non-target objects) and proportion of gaze following (saccades from face to target objects/ saccades from face to target and non-target objects). The results indicated that the same gaze-following behavior may involve different underlying mechanisms: the gaze-following in AD children may be volitional and inductive behaviors, while the gaze-following in TD may be reflexive and spontaneous behaviors. The volitional and inductive gaze-following suggested that the style of processing social information in children with AD were different compared with TD children. Children with AD could not flexibly distribute attention resource according to changes in social scene. They captured social information with the analytic way of feature process and lacked the understanding of the social meaning of social information.