Teaching Critical Thinking Skills in Science with sInvestigator

乔治梅森大学(GMU)一个由计算机科学、人工智能、科学教学和教育学、系统架构、软件工程、知识工程和人机交互领域的专家组成的团队正在开发一种认知辅助工具sInvestigator，以帮助学生在解决科学问题时获得批判性思维技能。SInvestigator认知助手是一项重大的技术创新，基于科学推理的计算理论。它结合了大量关于科学推理的常识和证据，指导和帮助学生进行科学探究的过程。该工具新颖，可在多个计算机平台上运行。它是为学生在家里和实验室中使用而设计的，并支持他们获得基于证据的推理的核心能力。由此产生的理论可以扩展到所有STEM学科，包括本科生和K-12。通过可容纳50-70人的讲习班扩大参与度。招聘的目标是社区学院和少数族裔服务机构，以吸引妇女和代表不足的群体参加与GMU教学和教师卓越中心合作举办的讲习班。基于证据的推理不仅在STEM学科中处于问题解决和决策的核心，而且在法律、情报分析、法医、医学、历史、考古学等领域也是如此。编写的教材和调查员练习广泛分发。教师和研究人员如何免费获得sInvestigator的信息发布在GMU的学习代理中心网站上。这个为期三年的项目首先在两个荣誉课程中试行，招收30-40名学生。在小组工作中，学生们被引导去处理一个科学问题，作为不断发现证据、假设和论点的过程。通过回答研究问题来增加新的知识，这些问题探索学生对科学过程技能的感知的改善，以及通过课程评估衡量的学生内容知识的增长。学生有很多机会发挥想象力和创造力，并获得批判性的科学实践，特别是：(1)提出问题；(2)构建解释；(3)从证据中进行辩论；以及(4)获得、评估和交流解释(NRC，2012 p.3)。SInvestigator的使用在一系列课程中进行了探索，以了解什么是可行的，什么是不可行的，并在开发和测试案例研究的同时逐步发展理论和方法。混合方法评估主要是形成性的，重点是通过调查和访谈了解学生与sInvestigator的合作体验。在评价的基础上对评价工具进行了修改，分析了学生的内在动机、职业动机、自我决定、自我效能感和年级动机。在最后一年，将使用准实验研究设计来评估sInvestigator在两个课程部分的250-300名学生参加的普通科学课程中的影响。该设计允许使用课程评估、科学动机问卷和科学过程技能清单对学生成绩进行比较。

A team of experts from computer science, artificial intelligence, science teaching and pedagogy, systems architecture, software engineering, knowledge engineering and human computer interaction at George Mason University (GMU) is developing a cognitive assistance tool, "sInvestigator," to help students acquire critical thinking skills in addressing scientific problems. The sInvestigator cognitive assistant is a significant technological innovation, based upon a computational theory of reasoning in science. It incorporates a substantial amount of general knowledge about scientific reasoning with evidence guiding and helping the student with the scientific inquiry process. The tool is novel and operates on multiple computer platforms. It is designed for students to use at home and in the lab and supports their acquisition of the core competency of evidence-based reasoning. The resulting theory is extendable to all STEM disciplines both undergraduate and K-12. Broadening participation is achieved through workshops having the capacity of 50-70 participants. Recruiting targets community colleges and minority serving institutions to attract women and underrepresented groups to the workshops organized in collaboration with the GMU Center for Teaching and Faculty Excellence. Evidence based reasoning is at the core of problem solving and decision making not only in STEM disciplines but also law, intelligence analysis, forensics, medicine, history, archeology and other domains. The developed educational materials together with sInvestigator exercises are widely distributed. Information on how teachers and researchers can freely obtain the sInvestigator are posted on GMU's Learning Agents Center website.The three-year project is first piloted in two Honors courses enrolling 30-40 students. Working in teams, students are guided to approach a scientific problem as ceaseless discovery of evidence, hypothesis and arguments. New knowledge is added by answering research questions that explore improvements in student perception of science process skills and gains in student content knowledge, as measured by course assessments. Students experience numerous opportunities to exercise imagination and creativity and acquire critical scientific practices, particularly: (1) Asking questions; (2) Constructing explanations; (3) Engaging in argument from evidence; and (4) Obtaining, evaluating and communicating explanations (NRC, 2012 p.3). The use of sInvestigator is explored in a sequence of courses to learn what works and what does not work and incrementally evolve the theory and the approach while developing and testing case studies. The mixed-methods evaluation is mostly formative with a focus on understanding students' collaborative experiences with sInvestigator through surveys and interviews. Modification of assessment tools is made based upon evaluation which allows for analysis of student intrinsic motivation, career motivation, self-determination, self-efficacy and grade motivation. In the final year a quasi-experimental research design will be used to evaluate the impact of sInvestigator in a general science course with 250-300 students in two course sections. The design allows for a comparison group on student achievement scores using course assessments, a Science Motivation Questionnaire and a Science Process skills inventory.