科学发现学习的认知模型建构

Cognitive Modeling of Scientific Discovery Learning

当前，利用虚拟学习环境进行科学发现学习已经成为推进科学教育的一条重要途径。然而，由于科学发现学习的认知过程高度复杂，而现有的虚拟学习环境未能提供所需的学习支持，导致其应用效果难以令人满意。要解决这个问题，就必须建立科学发现学习的认知模型，进而以此为基础进行虚拟学习环境的设计开发。因此，该文在科学哲学、信息加工心理学和教育科学相关研究的基础上，从领域知识的结构、发现机制和学习活动的设计三个方面建立科学发现学习的认知模型。科学发现学习是学生针对实际系统设计实验从而渐进地探究发现科学理论模型的过程。其中，科学理论模型包括局部模型与全局模型两个层面，和数学模型与物理模型两种形式。学生可以采用假设驱动或实验驱动两种策略在假设空间和实验空间中进行双重搜索，其知识形成过程可以分为从假设空间到局部模型和从局部模型到全局模型两个阶段。科学发现学习需经历理解问题、提出假设、设计实验、检验假设和得出结论五个环节，应从领域知识、元策略和内部目标三个方面为学生提供支持。该认知模型结构化特征明显，易于通过计算机代码进行形式化表达，为面向科学发现学习的智能化虚拟学习环境的设计开发奠定了必要的理论基础。

Presently， the use of virtual learning environment for scientific discovery learning has become an important way to implement science education. However， because of the highly complex cognitive process of scientific discovery learning and the lack of necessary learning support， the existing virtual learning environment failed to achieve satisfactory effect. To solve this problem， it is necessary to establish a cognitive model of scientific discovery learning as the basis for the design and development of virtual learning environment. Therefore， based on the philosophy of science， information processing psychology and education sciences research， this paper has built the cognitive model of scientific discovery learning from the structure of domain knowledge， the information process mechanism of discovery and the design of learning activity aspects. In scientific discovery learning， students gradually design experiment to explore scientific theory model of actual system. The scientific theory model includes partial model and complete model which may be described in mathematical or physical way. Scientific discovery learning is a dual search in learner's hypothesis space and experiment space guided by hypothesis-driven strategy or experiment-driven strategy. The knowledge construction of learner can be divided into two phases， one is from hypothesis space to partial models and the other is from partial models to complete model. Scientific discovery learning includes five steps: analyzing problem， raising hypothesis， designing experiment， testing hypothesis and drawing conclusion. Learning supports should be provided in domain knowledge， meta-strategy and internal target aspects.The structural characteristic of the cognitive model is obvious and easy to be formalized through computer code. Therefore， the cognitive model lays the theoretical foundation for the design and development of intelligent scientific discovery learning-oriented virtual learning environment.