📚 Edexcel & OCR A Level Physics June 2023 Paper 3: Practical and Synoptic Skills | Edexcel 与 OCR A Level 物理 2023年6月 Paper 3：实验与综合技能
This revision guide is written for candidates preparing for the June 2023 A Level Physics Paper 3. It brings together the practical, data-analysis and synoptic skills that appear in both the Edexcel and OCR specifications, with particular attention to the style of reasoning examiners actually reward in Paper 3. The central message is simple: mastering the underlying models, the graph work and the uncertainty calculations is far more effective than memorising isolated facts, because Paper 3 almost never asks a question you have seen before in exactly that form.
本复习指南专为准备 2023 年 6 月 A Level 物理 Paper 3 的考生编写。它汇集了 Edexcel 与 OCR 两个考试大纲中都会出现的实验、数据分析与综合应用技能，并特别关注考官在 Paper 3 中真正给分的推理方式。核心观点很简单：掌握基本模型、图像处理和不确定度计算，远比死记硬背孤立知识点更有效，因为 Paper 3 几乎不会以完全相同的形式重复你见过的题目。
Throughout this article, every principle is followed by a worked-style application so that you can see how a mark scheme converts a physical idea into marks. Read it with your specification, your formula sheet and your lab book open beside you.
在本文中，每一条原理后面都紧跟一个"应用示范"，让你看到评分标准是如何把一个物理概念转化为分数的。阅读时请把考试大纲、公式表和实验记录本放在手边。
1. Understanding the Paper 3 Format | 了解 Paper 3 考试形式
Paper 3 is where examiners separate students who have memorised equations from students who can think like physicists. In Edexcel Physics, 9PH0/03 General and Practical Principles in Physics lasts 2 hours 30 minutes, carries 120 marks and is worth 40 per cent of the A Level. In OCR A Level Physics, H556/03 Unified Physics lasts 1 hour 30 minutes, carries 70 marks and is worth 26 per cent of the A Level. Both are terminal papers taken at the end of the two-year course.
Paper 3 是考官用来区分"背过公式的学生"和"能像物理学家一样思考的学生"的一场考试。在 Edexcel 物理中，9PH0/03《General and Practical Principles in Physics》（通用与实验原理）考试时长 2 小时 30 分钟，满分 120 分，占 A Level 总成绩的 40%。在 OCR A Level 物理中，H556/03《Unified Physics》（统一物理）考试时长 1 小时 30 分钟，满分 70 分，占 A Level 总成绩的 26%。两门都是两年课程结束后参加的终结性考试。
Both papers are synoptic: any topic from the whole two-year course may appear, and a single question can move between mechanics, fields, waves, electricity, thermal physics and quantum physics. Neither paper is a practical examination, yet both demand practical reasoning: experimental design, instrument choice, uncertainty treatment, graph work and evaluation of method. The practical endorsement (CPAC in Edexcel, PAG-based in OCR) is reported separately and does not contribute marks to Paper 3, but the skills it certifies are exactly what Paper 3 tests indirectly.
两份试卷都具有综合性质：两年课程中的任何主题都可能出现，同一个题目可以在力学、场、波、电学、热学和量子物理之间切换。两份试卷都不是实操考试，但都要求实验推理能力：实验设计、仪器选择、不确定度处理、图像处理和方法评估。实验技能认证（Edexcel 的 CPAC、OCR 基于 PAG 的认证）单独报告，不计入 Paper 3 的分数，但它认证的技能正是 Paper 3 间接考查的内容。
A typical Paper 3 sequence looks like this: a short unfamiliar scenario, a table of raw measurements, then a request to plot a graph, find a gradient with its unit, use that gradient to obtain a physical quantity, and finally evaluate whether the result supports a stated hypothesis. Notice that the physics content may be familiar while the context is deliberately new - that is the design intention, not an accident.
典型的 Paper 3 题目顺序是这样的：一个简短而陌生的情境、一张原始数据表，然后要求作图、求出带单位的斜率、用斜率得到某个物理量，最后评估结果是否支持题目给出的假设。请注意：物理内容可能是熟悉的，但情境是刻意陌生的——这是命题的设计意图，而不是意外。
2. Specification Differences: Edexcel 9PH0/03 vs OCR H556/03 | 两大考纲差异对比
The two specifications test the same physics, but they weight the skills differently, and knowing the difference changes how you allocate revision time. Edexcel Paper 3 is longer and heavier, so it can afford multi-stage questions that build on each other; a mistake in part (a) can therefore cascade. OCR Unified Physics is shorter, so questions tend to be more self-contained and the synoptic links are usually signposted inside a single stem.
两个考纲考查的物理内容相同，但对各项技能的权重不同，了解这种差异会改变你分配复习时间的方式。Edexcel Paper 3 更长、分值更重，因此可以容纳层层递进的多步骤题目；第 (a) 问的错误可能引发连锁失分。OCR《Unified Physics》更短，题目往往更独立，综合联系通常在同一题干内部就被明确提示出来。
| Feature | 项目 | Edexcel 9PH0/03 | Edexcel | OCR H556/03 | OCR |
| Duration | 时长 | 2 h 30 min | 1 h 30 min |
| Marks | 分值 | 120 | 70 |
| Weighting | 占比 | 40% of A Level | 26% of A Level |
| Practical framework | 实验框架 | 16 core practicals, CPAC | 12 Practical Activity Groups (PAGs) |
| Typical flavour | 题目风格 | Multi-stage data response; extended 6-mark evaluations | Shorter structured parts; explicit synoptic stems |
One practical consequence: Edexcel candidates must rehearse carrying an answer forward from one part to the next with full precision, while OCR candidates must rehearse switching topic quickly within a single question. Both groups should practise writing conclusions that name a numerical result, its uncertainty and the physical reason it does or does not agree with theory.
一个实际的结果是：Edexcel 考生必须反复练习"全精度地把答案带入下一问"，而 OCR 考生必须练习"在同一道题中快速切换主题"。两组考生都应练习写出这样的结论：给出数值结果、其不确定度，以及它与理论相符或不相符的物理原因。
3. Experimental Design and Planning | 实验设计与方案规划
Planning questions are marked against a predictable checklist. Identify your independent variable, your dependent variable and every variable you must control. State the range of the independent variable and the number of readings. Describe how each quantity is measured and with which instrument. Explain how you will reduce the dominant uncertainty. Finally, say what you will plot and what the gradient or intercept will tell you.
方案设计题的评分遵循一份可预测的清单：确定自变量、因变量以及所有必须控制的变量；说明自变量的取值范围和读数个数；描述每个量如何测量、用哪种仪器；解释如何减小主要不确定度；最后说明你将作什么图，以及斜率或截距能告诉你什么。
- Independent variable: the quantity you deliberately change, with a stated range. | 自变量：你主动改变的量，需给出取值范围。
- Dependent variable: the quantity you measure, with its instrument and resolution. | 因变量：你测量的量，需给出仪器及其分辨率。
- Control variables: everything else that could affect the result, with the method of control. | 控制变量：其他所有可能影响结果的量，以及控制方法。
- Repeats and averaging: at least three readings per point, then a mean; anomalous results rejected with a reason. | 重复与平均：每个点至少测三次取平均；异常值需给出理由后剔除。
- Safety and ethics: a brief, specific risk statement - hot surfaces, high voltages, lasers, falling masses. | 安全与伦理：简短而具体的风险说明——高温表面、高压、激光、下落重物。
ne through the origin with gradient 4π² / g, so g = 4π² / gradient. Writing that sentence in the plan shows the examiner you know why the experiment is being done, not merely how.
优秀的答案还会预先规划图像。如果理论预测单摆满足 T² = (4π² / g) l，那么以 T² 对 l 作图，应得到一条过原点的直线，斜率为 4π² / g，因此 g = 4π² / 斜率。在方案中写出这句话，会让考官看到你知道这个实验"为什么做"，而不只是"怎么做"。
T² = (4π² / g) × l ⇒ g = 4π² / gradient
4. Measurements, Instruments and Resolution | 测量、仪器与分辨率
Examiners expect you to choose an instrument whose resolution is appropriate to the quantity, and to quote the reading to that resolution. A metre rule reads to 1 mm, vernier calipers typically to 0.1 mm or 0.05 mm, a micrometer screw gauge to 0.01 mm, a top-pan balance to 0.01 g, and a protractor to 1°. A stopwatch has a display resolution of 0.01 s but a human reaction uncertainty of roughly 0.2 s to 0.3 s, and you should quote the reaction time rather than the display resolution when timing by hand.
考官期望你选择分辨率与所测量相匹配的仪器，并按该分辨率读数。米尺读到 1 mm，游标卡尺通常读到 0.1 mm 或 0.05 mm，螺旋测微器读到 0.01 mm，电子天平读到 0.01 g，量角器读到 1°。秒表显示分辨率为 0.01 s，但人的反应不确定度约为 0.2 s 至 0.3 s；手工计时时应引用反应时间，而不是显示分辨率。
Three words are routinely confused and are routinely examined. Accuracy is how close a reading is to the true value, and is limited by systematic error. Precision is how closely repeated readings agree with each other, and is limited by random error. Resolution is the smallest change an instrument can display. A precise set of readings can still be inaccurate - that combination is the signature of a zero error or a calibration fault.
有三个词经常被混淆，也经常被考到。准确度指读数与真值的接近程度，受系统误差限制；精密度指多次重复读数彼此的一致程度，受随机误差限制；分辨率指仪器能显示的最小变化。一组精密度很高的读数仍可能不准确——这种组合正是零点误差或校准故障的标志。
C is not a ratio scale so use K in gas calculations. Second, record raw readings as raw as possible and round only at the end - premature rounding is one of the largest silent sources of lost marks in graphical questions.
另外两个习惯也能拿分。第一，代入公式前一律换算为 SI 基本单位：cm → m、g → kg、mA → A；摄氏度不是比例标度，气体计算中要使用 K。第二，原始读数尽量保留原始精度，只在最后一步修约——过早修约是图像题中最大的"无声失分源"之一。
5. Uncertainties: Types, Combining and Propagation | 不确定度：类型、合成与传递
Uncertainty work is the single highest-yield skill in Paper 3. Random errors scatter readings about a mean and are reduced by repeating and averaging; systematic errors shift every reading in the same direction and are not reduced by averaging - they must be identified and corrected, for example by checking for a zero error or by calibrating against a known standard.
不确定度处理是 Paper 3 中投入产出比最高的单项技能。随机误差使读数围绕平均值散布，可通过重复测量取平均来减小；系统误差使所有读数朝同一方向偏移，取平均无法消除，必须找出并修正，例如检查零点误差或用已知标准进行校准。
For a single analogue reading, the usual convention is to quote plus or minus half the smallest division; for a digital display, many mark schemes accept plus or minus the smallest division, though some accept half. When repeating, the uncertainty in the mean is best estimated as half the range of the repeats, provided you have at least three readings. Always follow the convention that the question itself uses.
对于单次模拟读数，通常约定为"± 最小分度的一半"；对于数字显示，许多评分标准接受"± 一个最小分度"，也有接受"± 一半分度"的。重复测量时，若至少有三个读数，均值的不确定度最好按"重复值极差的一半"估计。请始终沿用题目本身使用的约定。
| Operation | 运算 | Rule | 规则 | Example | 举例 |
| Addition / subtraction | 加减 | Add absolute uncertainties | 绝对不确定度相加 | (5.0 ± 0.1) + (3.0 ± 0.2) = 8.0 ± 0.3 |
| Multiplication / division | 乘除 | Add percentage uncertainties | 百分比不确定度相加 | 2% and 3% give 5% in the product |
| Powers | 乘方 | Multiply the percentage uncertainty by the power | 百分比不确定度乘以指数 | 2% in r gives 4% in r² |
| Graph gradient | 图斜率 | Use the steepest and shallowest acceptable lines | 用最陡与最平的可接受直线 | Δgradient / gradient × 100% |
), but at A Level the simple addition of percentage uncertainties is normally accepted and is easier to justify in writing. Whatever you use, show the substituted numbers, not just the final percentage.
更严谨的处理使用平方和根公式 Δz / z = √((Δx / x)² + (Δy / y)²)，但在 A Level 阶段，通常接受"百分比不确定度简单相加"，因为它更容易用文字说明。无论用哪种方法，都要写出代入的数字，而不是只给最后的百分比。
percentage uncertainty = (Δx / x) × 100%
6. Graphs, Linearisation and Error Bars | 图像、线性化与误差棒
Graph questions follow a fixed routine. Choose sensible axes so the plotted region fills at least half of the grid. Label each axis with the quantity, the symbol and the unit. Plot points with a small cross or a dot with a circle. Draw a single straight line of best fit with a ruler, balancing points above and below. Add error bars only when asked. Then take the gradient from two well-separated points that lie on the line, never from two data points.
图像题遵循固定的流程：选择恰当的坐标轴，使作图区域至少占网格的一半；每个轴标注物理量、符号和单位；用小的叉号或圆点作图；用直尺画一条最佳拟合直线，使点上下均衡分布；只有在要求时才加误差棒；然后用线上两个相距较远的点求斜率，绝不能用两个原始数据点求斜率。
The most common instruction is "show that the graph is a straight line and determine the gradient". A line is only proportional if it passes through the origin; if the intercept is non-zero, say so and interpret it, for example as a zero error or an inherent offset in the apparatus. Units of the gradient are the units of the y-axis divided by the units of the x-axis, and a missing gradient unit is a standard dropped mark.
最常见的指令是"证明图像为直线并求出斜率"。只有当直线过原点时才能称为正比；若截距不为零，要说明并解释其含义，例如零点误差或仪器固有偏移。斜率的单位等于 y 轴单位除以 x 轴单位，漏写斜率单位是一个典型的失分点。
ry is y = A x ⁿ, then plotting log y against log x gives a gradient n and an intercept log
更多咨询请联系16621398022（同微信）
或截图后用「扫一扫 → 相册」选择本图