近年来，中国高考改革和课程改革在如火如荼地进行中。新修订的《普通高中数学课程标准（2017年版）》提出了包含数学建模素养在内的六大数学核心素养，并给出了基于数学核心素养的高考命题建议。高考作为一项大规模、高利害的常模参照性考试，对一线教学具有重要的引导作用。为了研究高考数学试题对数学建模素养的考查情况，本文选取2021—2023年含新高考Ⅰ卷、新高考Ⅱ卷、全国甲（理）卷、全国乙（理）卷、北京卷、天津卷在内的18套高考试卷，针对以下三个问题展开研究：(1)如何构建指向数学建模素养的高考试题分析框架？(2)2021—2023年高考数学试题对数学建模素养的考查有怎样的特点？(3)如何根据试题分析结论制定教学策略？

本研究首先阅读、梳理文献，对数学建模素养的相关概念进行界定；其次，基于已有对数学建模素养的维度和水平划分的研究，结合试题的具体情况，参照一线教师的访谈结果和8位专家教授的指导意见，构建指向数学建模素养的试题分析框架；接着，利用此框架对18套高考试卷中指向数学建模素养的试题进行划分、编码和量化，并对各水平进行赋值。量化过程中，采取小题计数法，以各维度考查水平的加权平均值，作为试题在该维度的总体水平值衡量试题对这一维度的考查情况；以试题在各维度考查各水平的题目数量，衡量试题对不同水平的考查情况；以试题所有维度的水平值之和作为该试题的综合水平值，衡量试题对数学建模素养的综合考查水平。最后，对量化结果进行分析和讨论，得到以下结论：

第一，从维度和水平两个层面构建指向数学建模素养的高考试题分析框架。维度包括三方面：情境类型、思维层次、建模层次，每个维度均包括三个水平，其中，思维层次和建模层次均包含二级指标，分别为知识主题和建模表现。

第二，近三年高考试题对数学建模素养的综合考查水平呈现整体上升的态势，具体表现有：(1)试题对情境类型的考查主要集中在水平一和水平二；对思维层次的考查主要集中在水平一和水平二，其中，对知识主题的考查主要集中在“概率与统计”；对建模层次的考查主要集中在水平二，其中，对建模表现的考查主要集中在“分析和解决问题”。(2)试题对数学建模素养的综合考查处于中等水平，但整体难度在逐渐加大，对学生数学建模素养的要求也越来越高。

第三，指向数学建模素养提升的教学策略主要包括以下三方面：(1)通过阅读与分析理解情境，提高语言转化能力。(2)借助问题链推动解题教学，培养逻辑思维能力。(3)开展数学建模活动，培养发现问题和提出问题的能力。

In recent years, the reforms of the China College Entrance Examination(CEE) and curriculum have been progressing vigorously. The revised "Mathematics Curriculum Standards for Ordinary High Schools (2017 Edition)" proposes six mathematical core literacy areas, including mathematical modeling literacy, and provides suggestions for CEE proposition based on these core literacies. As a large-scale, high-stakes norm-referenced examination, the CEE plays an important guiding role in frontline teaching. To investigate the examination of mathematical modeling literacy in CEE mathematics test questions, this study selected 18 sets of CEE test papers from 2021 to 2023, including the New College Entrance Examination Volume I, New College Entrance Examination Volume II, National Type A (Science) Volume, National Type B (Science) Volume, Beijing Volume, and Tianjin Volume. The study focused on three questions: (1) How to construct an analytical framework for CEE test questions targeting mathematical modeling literacy? (2) What are the characteristics of the examination of mathematical modeling literacy in the 2021—2023 CEE mathematics test questions? (3) How to develop teaching strategies based on the conclusions drawn from the analysis of the test questions?

The research process involved an initial literature review and conceptual clarification of mathematical modeling competency. Subsequently, drawing on existing research on the dimensions and levels of mathematical modeling competency, combined with the specific characteristics of the test questions, insights from interviews with frontline teachers, and guidance from eight expert professors, an analytical framework for CEE test questions targeting mathematical modeling competency was developed. This framework facilitated the categorization, coding, and quantification of test questions related to mathematical modeling competency across the 18 sets of CEE test papers, allowing for the assignment of values to each level. Quantification involved employing a sub-question counting method and calculating weighted average values for each dimension to measure the overall level of assessment for a particular dimension. The number of questions assessing each level within each dimension was used to evaluate the examination of different competency levels. Furthermore, the sum of level values across all dimensions yielded a comprehensive level value for each test question, serving as an indicator of the overall assessment level of mathematical modeling competency. Finally, the quantified results were analyzed and discussed, leading to the following conclusions:

Firstly, an analytical framework for CEE test questions targeting mathematical modeling literacy was constructed at two aspects: dimensions and levels. The dimensions include three aspects: scenario types, thinking levels, and modeling levels. Each dimension consists of three levels, with the thinking levels and modeling levels including knowledge topics and modeling performance as secondary indicators.

Secondly, the comprehensive examination level of mathematical modeling literacy in the CEE test questions over the past three years showed an overall upward trend. Specifically: (1) The examination of scenario types in the test questions mainly focused on level one and level two, as well as the examination of thinking levels mainly focused on level one and level two, in which the examination of the knowledge subject mainly focuses on the "probability and statistics". As for the examination of modeling levels, which primarily focused on level two, has a particular emphasis on the examination of modeling performance in the contex of "analysis and problem solving". (2) The comprehensive examination of mathematical modeling literacy in the CEE test questions over the past three years was at a moderate level, but the overall difficulty has gradually increased, indicating higher requirements for students' mathematical modeling literacy.

Lastly, teaching strategies for enhancing mathematical modeling literacy mainly include the following three aspects: (1) Improving language translation skills by reading and analyzing scenario descriptions. (2) Promoting logical thinking abilities by using problem chains to teach questions. (3) Conducting mathematical modeling activities to cultivate the ability to identify and propose problems.