假装游戏是学龄前儿童典型的游戏形式，其出现标志着个体表征能力发展的飞跃，对儿童多领域的发展具有重要意义。然而，相比于普通儿童，孤独症儿童在假装游戏领域却表现出明显的障碍，这也成为了孤独症儿童主要的临床特征与早期诊断依据之一。国内外研究者对孤独症儿童假装游戏技能的干预进行了探究，其中，视频示范的干预策略颇受关注且应用最广。但在已有研究中，研究者们多聚焦于假装游戏中的功能性假装游戏和赋予属性等技能，而对虚构客体游戏技能的关注则较为缺乏。因此，本研究聚焦于假装游戏中的虚构客体游戏技能，旨在运用视频示范策略来提升孤独症儿童游戏技能的发展水平。 本研究运用单一被试研究中的跨行为多探测设计，以评价视频示范策略对孤独症儿童习得、维持并泛化虚构客体游戏技能的干预效果。干预过程共分为两个阶段，阶段一中，儿童以自身为游戏中介，做出指向自己的虚构客体游戏行为，研究者依次对“假装洗脸”、“假装吃饭”和“假装打扫卫生”三个目标活动介入干预；阶段二中，研究者引入了玩偶作为游戏的消极中介，并选择“假装给娃娃理发”、“假装照顾生病的娃娃”和“假装给娃娃洗澡”作为目标活动，儿童在与玩偶的互动中完成虚构客体游戏。各阶段的干预都包括基线期、干预期和维持泛化期，此外，研究者还在两阶段的研究中各加入了三个泛化活动，通过对比儿童在基线期和泛化期的表现来衡量干预的泛化效果。 研究表明，视频示范策略对孤独症儿童虚构客体游戏技能的习得、维持与泛化均具有良好的干预效果。经过干预后，孤独症儿童在两个阶段的目标游戏中表现出与视频脚本一致的游戏行为数量均有显著增加,且能够得到长期维持，儿童习得的虚构客体游戏技能也能够得到泛化，在未教学的新异游戏情境中能够表现出恰当的虚构客体游戏行为。

Pretend play is a typical form of play for preschool children. The emergence of pretend play marks a leap in the development of individual representational ability, which is of great significance to children's multi-field development. However, children with autism show obvious obstacles in the field of pretend play, which has become one of the main clinical features and early diagnosis basis of autism. Researchers have explored the intervention of children with autism in pretend play skills, among which the intervention strategy of video modeling has attracted much attention and been applied widely. Researchers have focused on the skills of functional play with pretense and assignment absent attributes, while the attention to the imagining absent objects play skills is relatively lacking in previous studies. Therefore, this study focuses on the imagining absent objects play skills, aiming to enhance the development level of play skills of children with autism by using video modeling. A single-case experimental design using a multiple probe across behaviors was used to examine the relationship between video modeling and the acquisition, maintenance, and generalization of imagining absent objects play skills. The intervention process is divided into two stages. Goals in stage 1 intervention involved self as agent; goals in stage 2 intervention involved doll as passive agent. Stage 1 instructional goals included pretending to face washing, eat a meal, and clean up. Stage 2 instructional goals were giving a haircut to a doll, taking care of a sick doll, and bathing a doll. Each stage of intervention includes baseline, intervention, maintenance and generalization sessions. In addition, three generalization activities were added to the two stages of the study. The generalization effect of intervention was measured by comparing the performance of children in baseline and generalization sessions. The results indicated that video modeling is effective in establishing and maintaining imagining absent objects play in all three children. Both children acquired scripted play behaviors quickly and maintained this performance during follow-up in two stages of intervention. Generalization to untaught imagining absent objects play activities occurred in three children, and they can show appropriate play behaviors.