【教师说】从IGCSE到A-Level：不是灌输公式，而是建构思维方式 I From IGCSE to A-Level: Constructing a Mindset for Science

作为一名国际化课程体系下的理科老师，我常常被问到这样一个问题：“理科到底该怎么学？”

是刷题？是背定义？还是记住每一个实验步骤？

我的答案是：都不是。

在我看来，从IGCSE到A-Level的这段旅程，本质不是知识的简单累加，不是公式的灌输，而是一次思维方式的系统升级。理科教育的真正目标，是让学生拥有面对未知世界的底层能力——观察、假设、验证、建模、反思。

兴趣不是被“表演”出来的。真正能点燃一个学生好奇心的，是一个让他无法用已有知识解释的现象。

在理科课堂上，我常用“反常”开场。比如讲气体定律时，我不会先给出PV=nRT公示，而是拿一个注射器堵住出口，让学生亲手推活塞：“你用力推，感觉越来越费力，但里面的空气质量变了吗？没有。那到底是什么在抵抗你？” 学生的固有认知被轻轻撼动——他们以为用力是因为空气被“压缩”了，但当被追问“压缩到底改变了什么”时，他们开始意识到：看不见的气体粒子间距缩小、分子碰撞频率提升才是核心原因。

兴趣源于困惑，困惑源于认知冲突。教师的任务不是立刻消除困惑，而是把困惑转化成可探究的问题。当一个学生说出“这好像和我想的不太一样”时，教育的契机就出现了。从IGCSE到A-Level，这种能力需要循序渐进深化：IGCSE阶段，我们更多用直观的反常现象制造冲突；A-Level阶段，则可以用更抽象的悖论（如“为什么理想模型和真实气体有偏差”）来持续维持学生的探究张力。

国际化课程非常强调探究能力，但真正的探究不是“实验步骤写得很清楚，学生照着做就行”。它的核心是：引导学生走完从提出问题、寻找证据到归纳结论的完整科研流程。

这个阶段，学生需要掌握最基本的科学方法：明确自变量与因变量、规范控制变量、重复测量取平均值、甄别实验异常数据。以“温度对反应速率的影响”实验为例，学生会提前收到一个半结构化的实验设计模板，但要求他们自己选择温度区间、确定测量方案、预判可能出现的误差。这个阶段的重点是规范习惯养成——学会用数据说话，而不是凭感觉下结论。

到了A-Level，学生需要面对更复杂、更开放的问题。如课题 “测定某种碳酸盐中的金属种类？”，我不给实验步骤，只给器材清单和参考文献提示。学生需要自主筛选滴定法 / 高温灼烧法等实验路径、评估方案优劣、核算实验不确定度，并在课堂研讨中客观说明实验结果的局限性。

这个过程中，最珍贵的是学生面对“不完美数据”时的态度。比如，在实验中产生的outlier数据，让学生自己做error analysis，仔细分析每一处可能的误差来源——滴定管读数精度、指示剂变色点的判断、甚至实验室温度对溶液体积的影响。最后他们基于探究实践，搞清楚了系统误差和随机误差的区别，对于科学现象和科学规律，不再出现绝对化的描述，而是懂得了要加限制条件和描述范围。这种对不确定性的坦然表达，恰恰是科学素养的核心，也是A-Level课程中高阶思维的直接体现。

从IGCSE到A-Level，理科思维的训练不是线性的知识积累，而是认知工具的逐级升级。我重点训练以下四种能力。

这是理科思维的起点。无论是物理实验还是化学反应，学生必须学会问：“哪些因素可能影响结果？如何逐一检验？”A-Level课程中的“变量拆解训练”，就是在一个复杂现象中（比如“化学反应速率的影响因素”），让学生列出所有可能因素，然后设计实验来区分相关性与因果性。这种训练在IGCSE阶段侧重识别变量，在A-Level阶段则进阶到多变量分析与交互作用。

理科的核心是从纷繁复杂的现实中提取本质。IGCSE阶段，学生学会用简单的模型解释现象（如粒子模型解释物态变化）；A-Level阶段，他们要理解模型的边界——理想气体定律在高压下为什么失效？点电荷模型在什么距离内近似成立？知道一个模型“不能解释什么”，往往比知道它能解释什么更重要。

一个训练有素的理科生，拿到一个数字会本能地问：“单位是什么？数量级合理吗？” 我会在新课开始前做“5分钟估算训练”：一个成年人的体积大约是多少升？一辆汽车以60km/h行驶，需要多长距离才能刹停？这种训练看似简单，但能有效防止学生陷入“计算出离谱答案却不自知”的困境。

在每个重要话题结束后，我会留出时间让学生反思自己的学习过程：“我最初是如何理解这个概念的？是什么改变了我的理解？哪种解释方式对我最有效——图表、数学推导，还是类比？”这种对自身思维过程的反思，不仅是科学学习的核心能力，也是终身学习的基础。

As a science teacher in an international curriculum system, I am often asked: “What is the right way to study science?”

Should I drill problem sets? Memorise definitions? Learn every experimental step by heart?

My answer is: none of the above.

In my view, the journey from IGCSE to A-Level is not about the simple accumulation of knowledge, nor the rote transmission of formulas. Rather, it is a systematic upgrade of thinking. The true purpose of science education is to equip students with the foundational abilities to face an unknown world: observation, hypothesis, testing, modelling, and reflection.

Interest is not something to be performed. What truly sparks a student’s curiosity is a phenomenon they cannot explain with their existing knowledge.

In my science classroom, I often begin with the unexpected. Take the gas laws: instead of presenting the formula PV=nRT, I hand a student a syringe, ask them to seal the outlet and push the plunger. “You feel increasing resistance,” I say. “But has the mass of air inside changed? No. So what is actually resisting you?”

The student’s existing intuition is gently unsettled. They may think the force comes from the air being “compressed.” But when pressed further - “What does compression really change?” - they begin to realise: what matters is the reduced distance between invisible gas particles and the increased frequency of their collisions with the walls.

Interest grows from puzzlement, and puzzlement from cognitive conflict. The teacher’s task is not to eliminate the confusion immediately, but to transform it into a question worth investigating. When a student says, “This isn’t quite what I thought,” a true teaching moment has arrived. From IGCSE to A-Level, this capacity needs to deepen progressively. At IGCSE, we often use striking, counter-intuitive phenomena to provoke conflict. At A-Level, we turn to more abstract paradoxes - such as why the ideal gas law deviates from the behaviour of real gases - to sustain the students’ drive for inquiry.

International curricula place a strong emphasis on inquiry skills, but genuine inquiry is not about following well-written lab steps. Its essence is guiding students through the complete cycle of scientific research: posing questions, gathering evidence, and drawing conclusions.

At this stage, students need to master the most basic scientific methods: identifying independent and dependent variables, controlling variables properly, taking repeated measurements and averaging them, and spotting anomalous data. In an experiment on “the effect of temperature on reaction rate,” students receive a semistructured design template - but they are required to choose the temperature range themselves, decide on a measurement scheme, and anticipate possible sources of error. The focus here is on developing disciplined habits: learning to speak with data, not with intuition.

At A-Level, students face more complex and openended problems. For example, in the task “Identify the metal in a given carbonate,” I provide no step-by-step procedure - only a list of equipment and references. Students must choose independently between different experimental approaches (such as titration or combustion), evaluate the strengths and weaknesses of each, calculate measurement uncertainty, and openly discuss the limitations of their results in class.

What is most valuable here is the students’ attitude toward imperfect data. When outliers appear, they conduct their own error analysis, carefully examining every possible source: the precision of the burette reading, the judgement of the indicator’s colour change, even the effect of room temperature on solution volume. Ultimately, they learn to distinguish between systematic and random errors. No longer do they describe scientific phenomena or laws in absolute terms; instead, they learn to attach limiting conditions and specify the scope of their conclusions. This candid acceptance of uncertainty lies at the very heart of scientific literacy, and it is a direct expression of higherorder thinking in the A-Level curriculum.

From IGCSE to A-Level, the training of scientific thinking is not a linear accumulation of knowledge, but a progressive upgrading of cognitive tools. I focus on developing the following four abilities.

This is the starting point of scientific thinking. Whether in a physics experiment or a chemical reaction, students must learn to ask: “Which factors might affect the outcome? And how can we test each one systematically?” In A-Level, “variable decomposition training” involves taking a complex phenomenon - such as the factors affecting reaction rate - listing all possible variables, and designing experiments to distinguish correlation from causation. At IGCSE, the emphasis is on identifying variables; at A-Level, it advances to multivariate analysis and interaction effects.

The essence of science is extracting what is essential from the complexity of reality. At IGCSE, students learn to explain phenomena using simple models - for example, the particle model for changes of state. At A-Level, they must understand the boundaries of those models: why does the ideal gas law break down at high pressure? Within what distance does the pointcharge model remain a good approximation? Knowing what a model cannot explain is often more important than knowing what it can.

A well-trained science student, when seeing a number, instinctively asks: “What are the units? Is the magnitude reasonable?”
At the start of a lesson, I often do a “five-minute estimation exercise”: What is the approximate volume of an adult, in litres? If a car is travelling at 60 km/h, how far does it need to come to a complete stop? These exercises may seem simple, but they effectively prevent students from arriving at absurd answers without realising it.

After each major topic, I set aside time for students to reflect on their own learning process: “How did I first understand this concept? What made my understanding change? Which mode of explanation worked best for me - graphs, mathematical derivation, analogy?” This reflection - thinking about one’s own thinking - is not only a core ability in science learning, but also the foundation of lifelong learning.