STEM教育强调将原本分散的四门学科知识自然组合形成整体，并用于解决真实世界中的问题。其宗旨是整合创新，应具备问题驱动、创中学、协作探究和设计制造等外部特征。在当前“互联网+教育”的大背景下，打破时空界限进行学习的需求愈发旺盛。这也对STEM教育提出一个新需求：从完全的线下转向线下线上相结合。我们将这种新需求界定为“异地同步”。但是对于诸如STEM教育这类需要创中学、需要协作探究、需要设计制造的课程来说，异地同步的学习空间的设计所面对的是一个需求含混的复杂教育系统。目前相关研究鲜有对STEM领域学习空间的设计研究，对异地同步的STEM学习空间的研究则更少。基于此，本研究聚焦于异地同步的STEM学习空间的设计研究，结合具体的STEM课程产品和学习空间概念模型构建，通过设计开发、教学实施、缺陷分析等过程，开展个案研究，探索适应异地同步的条件、满足STEM课程产品实际需求的STEM学习空间的设计方法。

本研究开发了《让书签发光：设计投影仪》STEM课程产品，基于文献研究和课程产品的实情初步构建了异地同步的STEM学习空间概念模型，包含物理空间、信息空间和交往空间三个子空间，尝试构建了异地同步的STEM学习空间设计方法原型（包含开发/确定课程产品、明确角色行为和一级需求、分析二级需求、设计学习空间、激活学习空间和一致性分析及优化六个步骤），并按照该设计方法实施了个案研究，结合技术人造物缺陷分析法，利用角色行为编码和教学过程机制图分析教学过程中的异常行为，回溯概念模型、方法原型和课程产品的问题并进行了修正。通过比较分析两轮的实施态数据与设计态数据，本研究得出结论如下：异地同步的STEM学习空间的概念模型和设计方法可以用于在异地同步条件下STEM学习空间的设计，它将有助于“互联网+STEM”的实现，但是该方法在分析二级需求步骤中所使用的需求解释表中的“环境最低噪音”需求属于默认或前置条件，可剔除。

本研究的不足之处：实施时的课程安排较为紧凑，且全手工分析过于繁琐、费力，容易出错返工，因此迭代次数少，可能无法暴露出全部的异常行为，可能仍有一些设计需求没有被鉴别出来。

STEM education emphasizes the natural combination of the originally scattered knowledge of four disciplines to form a whole, and it is used to solve problems in the real world. The aims of integration and innovation should equip with STEM with external characteristics including problem-driven, creating-while-learning, collaborative analysis and design and manufacturing. Given the current background of "Internet+Education", the growing requirement of breaking time-space boundary for learning seems inevitable. In terms of STEM education, it should respond to this new requirement by the shift from complete offline teaching to the integrated approach of online and offline teaching. We define this new requirement as “Synchronization in Different Places”. However, due to the external characteristics, which are listed previously, for courses such as STEM education, they are facing a complex education system with ambiguous requirements when it comes to the design of integrated learning space. At present, there is little research on the design of STEM learning space, and even less on the STEM learning space synchronized in different places. Therefore, this study focuses on the design of STEM learning space synchronized in different places, with the combination of specific STEM curriculum products and conceptual model construction of learning space. Through the processess including design and development, teaching implementation, defect analysis, etc., particular case studies are carried out, consequently resulting in the exploration of the prerequisites. Ultimately, this design method for STEM learning space may meet the actual needs of STEM curriculum products.

This research develops the STEM course product "Let Bookmarks Shine: Design Projector". Based on the literature research and the facts of the course product, this research constructs an initial conceptual model of STEM learning space in different places, including three subspaces: Physical Space, Cyber Space and Communication Space. This paper attempts to construct a prototype of synchronous STEM learning space design method in different places (including six steps: developing/determining curriculum products, defining role behaviors and first-level requirements, analyzing second-level requirements, designing learning space, activating learning space and consistency analysis and optimization). According to this design method, a case study is carried out. Combined with the technical artifact defect analysis method, the abnormal behaviors in the teaching process are analyzed by using role behavior coding and teaching process mechanism diagram. The problems of conceptual model, method prototype and curriculum products are traced back and corrected. By comparing and analyzing the two rounds of implementation data and design data, this study draws the following conclusions: the conceptual model and design method of STEM learning space synchronized in different places can be used in the design of STEM learning space synchronized in different places, which will contribute to the realization of "Internet + STEM". However, the requirement of "minimum environmental noise" in the requirement explanation table used in the second-level requirement analysis step of this method belongs to the default or pre-condition and can be eliminated.

The limitations of this study are as follows: the curriculum arrangement is relatively compact in terms of the implementation part; due to the complication of the manual analysis, mistakes could be easily made. Therefore, these two limitations may lead to the few rounds of iterations, fundamentally resulting in the little possibility that not all abnormal behaviors could be exposed, and some design requirements may still be unidentified.