通过文献研究法、访谈法和问卷调查法等，本研究构建了一个适用于计算机互动测评的科学探究能力测评框架，该框架把科学探究能力划分为4个要素并界定了4个要素的内涵，并把各要素划分为高、中、低3个表现水平，编制出各要素的可观测变量及编码，然后开发了基于计算机的“探索微藻的奥秘”互动测评工具并开展效化研究。研究结果显示：

（1）本研究开发的科学探究能力测评工具有良好的效度和信度，其中内容效度S-CVI/Ave为0.97，克隆巴赫α系数为0.74，Rasch模型分析得到被试信度为0.74，测试题的信度为0.99，说明学生群体能力分布具有一定代表性，题目的一致性程度较高。

（2）根据测评框架和测评工具构建科学探究能力与可观测变量的贝叶斯网络模型，能对被试的科学探究能力水平、各构成要素的水平进行有效地推断，推断结果能帮助教师开展针对性教学。

（3）与纸笔测验相比，该测评工具更加接近真实的科学探究，能有效激发学生的科学探究思维，学生在进行互动探究测试有自主学习的发生，测评系统的互动反馈特性能使学生获得了新的知识和思考。

In the 21st century, major international organizations and countries have proposed core literacy frameworks and indicators for student development in the 21st century, and these frameworks and indicators consistently include "active inquiry" as a core competency. There is a trend of shifting from paper-and-pencil tests to computer-based assessments. In China's High School Curriculum (2017 Edition), the core literacies of physics, chemistry, and biology all include "scientific inquiry". In order to implement the requirements of the curriculum in high schools, qualified scientific inquiry assessment tools need to be developed to measure the level of students' scientific inquiry literacy development and provide feedback to teaching. At the same time, due to the limitations of biology inquiry experiments and paper-and-pencil tests, it is urgent and important to develop a computer-based interactive inquiry assessment tool. To this end, this study aims to build a framework for assessing scientific inquiry skills by combining interactive computer-based assessment methods, designing and developing a computer-based assessment tool with interactive capabilities, validity and reliability, and determining the performance characteristics and levels of students’ scientific inquiry skills.

Through literature research, interviews, and questionnaires, this study constructed a framework for assessing scientific inquiry skills for computer-based interactive assessment. Then, a computer-based interactive assessment tool "Exploring the Mystery of Microalgae" was developed and validated. The results of the study showed that.

(1) The scientific inquiry ability assessment tool developed in this study has good validity and reliability, including content validity S-CVI/Ave of 0.97, Cronbach's alpha coefficient of 0.74, Rasch model analysis to obtain subject reliability of 0.74, and test question reliability of 0.99, indicating that the ability distribution of the student group is representative and the degree of consistency of the questions is high.

(2) The Bayesian network model of scientific inquiry ability and observable variables was constructed based on the assessment framework and assessment tools, which can effectively infer the level of scientific inquiry ability of the subjects and the level of each component, and the inferred results can help teachers carry out targeted teaching.

(3) Compared with paper-and-pencil tests, the assessment tool is closer to real scientific inquiry and can effectively stimulate students' scientific inquiry thinking, students have autonomous learning occurring in conducting interactive inquiry tests, and the interactive feedback feature of the assessment system enables students to gain new knowledge and thinking.