科学推理是高中生物学核心素养科学思维的重要组成部分,《普通高中生物学课程标准（2017年版2020年修订）》明确指出生物学对培养学生科学推理能力发挥着重要作用，评价高中生科学推理能力水平是一项基础性工作，但是目前广泛采用的科学推理能力测评工具主要以物理学为情境，生物学学科背景下的测评工具很少，本研究致力于开发质量合格的高中生物学科学推理能力测评工具。

首先，本研究以课程标准、经典的教育评价理论和现有的生物学核心素养测评框架为三大基石构建测评框架，最终构建出五维度四水平的高中生物学科学推理能力测评框架，包括相关性推理、概率推理、比例推理、控制变量推理、假设演绎推理五个子维度，在各个子维度下细分为 1、2、3、4 四个水平。

其次，依据项目反应理论的 Rasch 模型确定测评工具的项目类型为二阶选择题，试题数量为 20 道、计分方式为三级评分法、总分为 40 分，结合一线教师、学科专家和学生的建议编制测评工具；使用测评工具进行预测，收集到有效预测数据五十份，结合 Rasch 模型使用 WINSTEPS 软件分析测评工具质量，预测测评工具的区分度为 3.62、信度为 0.93和解释率为 70%；根据预测结果对题 8、题 11 和题 12 等项目进行修改，在预测测评工具的基础上加上生物成绩、学业总成绩、性别等基本信息以了解这些信息与科学推理能力的相关性，至此形成正式的高中生物学科学推理能力测评工具；正式测评工具的区分度为 9.08、信度为 0.99 和解释率为 85%，本研究最终开发出信度非常好、区分度良好、拟合度良好的高中生物学科学推理能力测评工具。

再者，使用测评工具对 169 名高三学生进行正式测试，了解到高三学生的科学推理能力总体水平较高，但整体上仍有进步空间；科学推理能力各个子维度的发展水平存在差异性，控制变量推理和假设演绎推理两个子维度发展水平最高，概率推理和比例推理发展水平次之，相关性推理的发展水平最差；科学推理能力与生物学成绩、学业总成绩存在 0.01 水平的显著正相关性，未发现存在显著性别差异。

本研究一方面基于高中生物学这一特定知识领域开发出质量合格的科学推理能力测评工具，有助于更准确地考查高中生生物学科学推理能力，另一方面获得了丰富的高中生生物学科学推理能力水平实证数据，为后续教育工作者们评价生物学科学推理能力提供数据支持。

Scientific reasoning is an important component of scientific thinking in the core literacy of high school biology, and China's "General High School Biology Curriculum Standards (2017 Edition Revised 2020)" clearly points out that biology plays an important role in cultivating students' scientific reasoning ability, and evaluating the level of scientific reasoning ability of high school students is a fundamental task, but currently the widely used scientific reasoning ability assessment tools are mainly contextualized in physics. This study is dedicated to developing a quality-qualified assessment tool for scientific reasoning ability in high school biology.

Firstly, this study uses curriculum standards, classical educational assessment theories and existing biology core literacy assessment frameworks as the three cornerstones to construct the assessment framework, and finally constructed a five-dimensional and four-level high school biology scientific reasoning ability assessment framework. It includes five sub-dimensions of conservation reasoning, probability reasoning, proportional reasoning, controlling variable reasoning, and deduction variable reasoning, which are subdivided into four levels of 1, 2, 3, and 4 under each sub-dimension.

Secondly, the Rasch model of item response theory is used to determine the item type of the assessment tool as second-order multiple-choice questions, the number of questions is 20, the scoring method was three-level scoring, and the total score is 40, and the assessment tool is prepared by combining the suggestions of front-line teachers, subject experts and students. Combined with the Rasch model, WINSTEPS software is used to analyze the quality of the assessment tool, and the discriminant of the assessment tool is 3.62, the reliability is 0.93, and the interpretation rate is 70%. Questions 8, 11, 12 and other items are modified according to the predicted results. Basic information such as biology test score, total academic score, gender and so on were added on the basis of the prediction assessment tool to understand the correlation between these information and scientific reasoning ability, so as to form a formal scientific reasoning ability assessment tool for high school biology. The formal high school biology scientific reasoning ability assessment tool has a discriminant of 9.08, a reliability of 0.99, and an interpretation rate of 85%. This study finally developed a high school biology scientific reasoning ability assessment instrument with very good reliability, good discriminant and good fit. There is a significant positive correlation between scientific reasoning ability and biology test score and total academic scores at the 0.01 level, and no significant gender differences is found。

On the one hand, based on the specific knowledge field of high school biology, this study develops a quality- qualified scientific reasoning ability assessment tool, which is helpful to more accurately examine the biology scientific reasoning ability of high school students. On the other hand, it obtains abundant empirical data on the level of biology scientific reasoning ability of high school students, which provides data support for subsequent educators to evaluate the biology scientific reasoning ability.  Scientific reasoning is an important component of scientific thinking in the core literacy of high school biology, and China's "General High School Biology Curriculum Standards (2017 Edition Revised 2020)" clearly points out that biology plays an important role in cultivating students' scientific reasoning ability, and evaluating the level of scientific reasoning ability of high school students is a fundamental task, but currently the widely used scientific reasoning ability assessment tools are mainly contextualized in physics. This study is dedicated to developing a quality-qualified assessment tool for scientific reasoning ability in high school biology. Firstly, this study uses curriculum standards, classical educational assessment theories and existing biology core literacy assessment frameworks as the three cornerstones to construct the assessment framework, and finally constructed a five-dimensional and four-level high school biology scientific reasoning ability assessment framework. It includes five sub-dimensions of conservation reasoning, probability reasoning, proportional reasoning, controlling variable reasoning, and deduction variable reasoning, which are subdivided into four levels of 1, 2, 3, and 4 under each sub-dimension. Secondly, the Rasch model of item response theory is used to determine the item type of the assessment tool as second-order multiple-choice questions, the number of questions is 20, the scoring method was three-level scoring, and the total score is 40, and the assessment tool is prepared by combining the suggestions of front-line teachers, subject experts and students. Combined with the Rasch model, WINSTEPS software is used to analyze the quality of the assessment tool, and the discriminant of the assessment tool is 3.62, the reliability is 0.93, and the interpretation rate is 70%. Questions 8, 11, 12 and other items are modified according to the predicted results. Basic information such as biology test score, total academic score, gender and so on were added on the basis of the prediction assessment tool to understand the correlation between these information and scientific reasoning ability, so as to form a formal scientific reasoning ability assessment tool for high school biology. The formal high school biology scientific reasoning ability assessment tool has a discriminant of 9.08, a reliability of 0.99, and an interpretation rate of 85%. This study finally developed a high school biology scientific reasoning ability assessment instrument with very good reliability, good discriminant and good fit. There is a significant positive correlation between scientific reasoning ability and biology test score and total academic scores at the 0.01 level, and no significant gender differences is found On the one hand, based on the specific knowledge field of high school biology, this study develops a quality- qualified scientific reasoning ability assessment tool, which is helpful to more accurately examine the biology scientific reasoning ability of high school students. On the other hand, it obtains abundant empirical data on the level of biology scientific reasoning ability of high school students, which provides data support for subsequent educators to evaluate the biology scientific reasoning ability.