📚 CIE A-Level Physics Practical: Exam Revision Essentials | CIE A-Level物理实验考前重点梳理
The CIE A-Level Physics practical examination (Paper 3) tests your ability to carry out experiments, record data accurately, analyse results, and evaluate procedures. This last-minute guide summarises the essential skills, common pitfalls, and key points you need to master for a high score.
CIE A-Level物理实验考试（Paper 3）考查你设计实验、准确记录数据、分析结果和评估实验方案的能力。这份考前冲刺指南梳理了必备技能、常见丢分点和关键得分技巧，帮助你高效备考。
1. Overview of the CIE Physics Practical Paper | 考试概览
The paper lasts 2 hours and usually consists of two questions, each worth 20 marks. One question may involve a familiar experiment, while the other could be an unfamiliar setup requiring you to apply your skills. You are expected to take measurements, record them in a table, calculate derived quantities, plot a graph, determine a gradient or intercept, and evaluate the experiment.
考试时长2小时，通常包含两道题，每题20分。一题可能为常见实验，另一题可能是陌生情境，需要灵活运用实验技能。你需要完成测量、记录数据、计算派生量、绘制图像、确定斜率或截距，并进行实验评估。
Marks are not awarded for "getting the right answer" alone. The examiner is looking for evidence of correct technique: sensible apparatus, a complete table with units, a well-drawn graph with a labelled line of best fit, a gradient with a sensible number of significant figures and an uncertainty, and an evaluation that links a stated problem to a specific improvement.
分数并不只看"答案对不对"。考官关注的是规范操作的证据：合理的器材选择、带单位的完整表格、标注最佳拟合线的规范图像、有效数字合理且带不确定度的斜率，以及把"问题"与"具体改进"联系起来的评估。理解评分逻辑，比盲目刷题更有效。
Paper 3 also rewards independence. If you simply copy a textbook method when the question asks you to "describe how you would determine X", you may lose the planning marks. Always respond to the exact wording of the question in front of you.
Paper 3同时奖励独立思考能力。如果题目要求"描述你如何测定X"，而你只是照搬课本方法，很可能丢掉设计分。务必针对眼前这道题的具体措辞作答，而不是背诵套路。
2. Time Management and Paper Strategy | 时间分配与答题策略
Time management is crucial: aim to spend about 25 minutes on measurements, 25 minutes on graph and analysis, and 10 minutes on evaluation per question. Always read the instructions carefully before touching any apparatus.
时间管理至关重要：每个问题建议约25分钟用于测量，25分钟用于图像与分析，10分钟用于评估。动手前一定要先把指令读清楚。
A practical exam rewards a calm, ordered routine. Spend the first three to five minutes reading both questions completely, noting the apparatus available, the variables named, and the number of marks for each part. Then start with the question you understand best so that you bank marks early.
实验考试奖励冷静有序的节奏。开考前三到五分钟把两道题完整读完，标出可用的器材、题目点名的变量以及每个小问的分值。然后从你最有把握的那道题开始，先把分数拿到手。
Never leave the table blank while you "think about the method". Take a trial reading first: it tells you whether the range of values will be wide enough to give a good graph. If the readings barely change, change the apparatus or the range before you commit to a full set.
不要一边空着表格一边"想方法"。先做一次试测：它能告诉你取值范围是否足以画出好图像。如果读数几乎不变，就应及时调整器材或量程，而不是硬着头皮测完全部数据。
Keep an eye on the clock at three checkpoints: after your first complete set of readings, after plotting the axes, and before you start the evaluation. If you are running late, prioritise the graph and gradient, because they carry many marks and take less time than extra repeats.
在三个时间点看表：完成第一组读数后、画完坐标轴后、开始写评估前。如果时间紧张，优先保证图像和斜率，因为它们分值高、耗时相对少，而重复测量的收益往往低于它们的代价。
3. Reading the Question and Planning the Experiment | 审题与实验设计
Every practical question contains hidden instructions. Words such as "state", "describe", "explain", "suggest" and "evaluate" demand different depths of answer. Underline the independent variable, the dependent variable and the quantities to be kept constant before you write anything.
每道实验题都藏着指令词。"state（陈述）""describe（描述）""explain（解释）""suggest（提出）""evaluate（评估）"要求的作答深度完全不同。动笔前先划出自变量、因变量以及需要保持不变的量。
When asked to describe a method, use a logical sequence: apparatus, what is measured, how it is measured, the range of values, and how many readings. A common structure is: set up, measure the independent variable, measure the dependent variable, repeat, then change and repeat again.
被要求描述方法时，按逻辑顺序写：用什么器材、测什么、怎么测、取值范围、测几组。一个通用结构是：搭建装置 → 改变并测量自变量 → 测量因变量 → 重复 → 再改变自变量并重复。
For a "suggest improvements" question, each improvement must include three elements: what to do, why it helps, and how it affects the result. An improvement without a reason usually scores zero.
"提出改进"类题目，每条改进都要包含三要素：做什么、为什么有用、对结果有什么影响。只写改进而不给理由，通常一分不得。
- Use a larger distance or angle to reduce the percentage uncertainty. | 增大距离或角度，以降低百分比不确定度。
- Repeat and average to reduce the effect of random error. | 重复测量取平均，以减小随机误差的影响。
- Use a fiducial marker or clamp to fix the position of the eye or apparatus. | 用标记线或夹具固定视线或器材位置。
- Use a more precise instrument, for example a digital vernier or a micrometer. | 换用更精密的仪器，例如数显游标卡尺或千分尺。
4. Accurate Measurement and Apparatus Choice | 准确测量与仪器选择
Choose the instrument that matches the size of the quantity you are measuring. A metre rule is fine for lengths of 20 cm to 1 m, a vernier caliper suits 1 cm to 15 cm with a precision of 0.01 cm, and a micrometer suits a few millimetres with a precision of 0.01 mm.
仪器要匹配被测量的大小。米尺适合20 cm至1 m的长度，游标卡尺适合1 cm至15 cm、精度0.01 cm，千分尺适合几毫米、精度0.01 mm。
h ±0.01 s for the reading but the human reaction time is about ±0.2 s to ±0.3 s. A digital ammeter usually has an uncertainty of ±1 in the last displayed digit, or ±half of it if the display is stable.
常见绝对不确定度可以这样写：米尺 ±0.5 mm 或 ±1 mm，游标卡尺 ±0.01 cm，千分尺 ±0.01 mm，数显秒表读数 ±0.01 s，但人的反应时间约 ±0.2 s 至 ±0.3 s。数显电流表通常取末位 ±1 个单位；若显示稳定，也可取末位的一半。
Watch out for parallax error. When reading a scale, place your eye directly level with the pointer, and for a liquid meniscus read the bottom of the curve. For a pendulum, time 10 or 20 oscillations and divide, which reduces the reaction-time uncertainty by the same factor.
注意视差误差。读刻度时视线要与指针齐平，读液体凹液面时读最低点。单摆实验要测10次或20次全振动的总时间再除以次数，这样反应时间带来的不确定度会按相同倍数缩小。
Where a quantity is zeroed, check the zero reading before use. A zero error in a micrometer or a balance must be recorded and either corrected or subtracted, otherwise every value carries the same systematic offset.
凡是可以"调零"的量，使用前都要检查零点。千分尺或天平的零点误差必须记录，并加以修正或扣除，否则每个数据都会带上同样的系统偏差。
5. Recording Data: Tables, Units and Significant Figures | 数据记录：表格、单位与有效数字
A correct table has quantity and unit in the header, raw readings on one line or column, and derived values clearly separated. Units must never be written next to every number inside the body of the table.
规范的表格应把"物理量+单位"写在表头，原始读数分行或分列记录，派生量单独成列并标注清楚。单位绝不能写在表格内容区的每一个数字后面。
| Quantity | l / cm | t₁ / s | t₂ / s | mean t / s | T / s |
| 1 | 50.0 | 14.2 | 14.4 | 14.3 | 0.715 |
| 2 | 60.0 | 15.6 | 15.7 | 15.65 | 0.783 |
Raw readings should all be given to the same number of decimal places, matching the precision of the instrument. Derived quantities should be given to a consistent number of significant figures, usually three, and never more than the data justify.
原始读数的十进制位数应一致，与仪器精度匹配。派生量取一致的有效数字，通常三位，且绝不超过原始数据所能支持的精密度。
s. Anomalous results should be identified and excluded from the mean, with a note in the table.
写不确定度时，不确定度取一位有效数字，并与测量值的小数位对齐。例如写 2.4 ± 0.2 s，而不是 2.43 ± 0.21 s。异常值应被标出并从平均值中剔除，同时在表格中注明。
6. Dealing with Uncertainties and Errors | 不确定度与误差处理
Percentage uncertainty converts every measurement to a common scale so you can compare their importance. It is calculated by dividing the absolute uncertainty by the measured value, then multiplying by 100.
百分比不确定度把所有测量换算到同一尺度上，便于比较它们的重要性。计算方法是用绝对不确定度除以测量值，再乘以100。
percentage uncertainty = (Δx ÷ x) × 100%
For combinations, the rules are simple. When quantities are added or subtracted, add the absolute uncertainties. When they are multiplied or divided, add the percentage uncertainties. When a quantity is raised to a power, multiply its percentage uncertainty by that power.
组合运算的规则很简单。加减法：绝对不确定度相加。乘除法：百分比不确定度相加。乘方：把该量的百分比不确定度乘以幂次。
Z = AⁿBᵐ ÷ Cᵖ → %U(Z) = n·%U(A) + m·%U(B) + p·%U(C)
Repeat readings mainly reduce random error, so the uncertainty in the mean can be taken as about half the range of the repeats. Systematic errors, such as a zero error or a wrongly calibrated scale, are not reduced by repeating; they must be identified and corrected.
重复测量主要减小随机误差，因此平均值的绝对不确定度可近似取重复读数极差的一半。系统误差（如零点误差、刻度未校准）不会因重复而减小，必须被识别并加以修正。
A useful exam habit is to identify which measurement contributes the largest percentage uncertainty and say so in your evaluation. Improvements should then target that quantity, because that is where the biggest gain lies.
一个很实用的应试习惯是：找出贡献百分比不确定度最大的那个量，并在评估中点明。改进措施就应该针对它，因为这里的收益最大。
7. Plotting Graphs Correctly | 图像绘制规范
Axes must be labelled with the quantity and the unit, for example "T² / s²" on the y-axis and "l / cm" on the x-axis. The scale should be chosen so that the plotted points occupy at least half of each axis, and it must be linear, using easy multiples such as 1, 2, 5 or 10 per major division.
坐标轴必须标注物理量和单位，例如纵轴写 "T² / s²"，横轴写 "l / cm"。刻度选择要让数据点至少占满每个坐标轴的一半，且必须是线性刻度，用1、2、5、10这类整数倍数作大格。
Plot points with a small sharp cross or a circle with a dot in the centre, so that the centre of the mark shows the true position. Do not use oversized dots, because they hide the position of the point and make the examiner's job harder.
数据点要用小而清晰的十字或"带点圆圈"标出，让标记中心代表真实位置。不要用很大的圆点，因为那会掩盖点的准确位置，也增加考官判读的难度。
Draw a single straight line of best fit that balances the points above and below it. In a practical exam you are usually expected to draw a straight line, so if the points curve, check whether you have chosen the wrong variables to plot.
画一条最佳拟合直线，使直线上下方的点大致平衡。实验考试通常期望得到直线；如果数据明显弯曲，先检查是不是坐标变量选错了。
For a curved relationship, do not draw a freehand curve through a scatter unless the question demands it. Instead, transform the variables so that the relationship becomes linear, then plot the transformed values.
若关系为曲线，除非题目明确要求，不要随手描一条平滑曲线。更好的做法是变换变量，使关系线性化，再绘制变换后的量。
8. Determining Gradient and Intercept | 斜率与截距的求解
The gradient must be found from the line of best fit, not from data points. Choose two points on the line, as far apart as possible, ideally more than half the length of the drawn line, and mark the triangle clearly on the graph.
斜率必须从最佳拟合直线上取点计算，不能直接用数据点。取直线上的两点，尽量相距最远，最好超过所画直线长度的一半，并在图上画出清晰的三角形。
Read each coordinate to the precision of half a small square and show your substitution in full. Writing the substitution makes your method visible, which is where the method marks lie.
每个坐标读到半小格的精度，并完整写出代入过程。写出代入式能让你的方法"可见"，方法分正是从这里来的。
gradient = (y₂ − y₁) ÷ (x₂ − x₁)
Do not round the coordinates before subtracting; rounding too early introduces avoidable error. Give the final gradient to two or three significant figures, consistent with the precision of your data, and always state the unit.
相减之前不要先对坐标四舍五入，过早取整会引入本可避免的误差。最终斜率取两到三位有效数字，与数据精度一致，并且一定要写单位。
The intercept is where the line crosses the axis, and it must be read from the graph, not from a data point. If the intercept is off-scale, extend the axis or the line, showing the extrapolated part as a dashed line. Quote the intercept with its unit, and remember that a negative intercept is meaningful information, not a mistake.
截距是直线与坐标轴的交点，必须从图上读出，而不是取数据点。若截距超出坐标范围，应延长坐标轴或直线，外推部分用虚线表示。截距要带单位；负截距往往是有意义的物理信息，而不是错误。
To estimate the uncertainty in the gradient, draw the steepest and shallowest acceptable lines through the extreme points and find the difference between the two gradients. The uncertainty is half of that difference.
估算斜率不确定度的方法是：通过极端数据点画出最陡和最平的两条可接受直线，求两斜率之差，不确定度取该差值的一半。
9. Linearising Relationships | 线性化处理
Many experiments give a power law or an exponential relationship that must be converted into a straight line before a gradient can be measured. The standard techniques are logarithms for power laws and natural logarithms for exponentials.
很多实验给出的是幂函数或指数关系，必须先转化为直线才能求斜率。标准做法是：幂函数取对数，指数关系取自然对数。
y = kxⁿ → lg y = n lg x + lg k
form, plotting lg y against lg x gives a gradient of n and an intercept of lg k, so k = 10intercept. The logarithm has no unit, so the axes are labelled "lg (T / s)" or similar, with the quantity divided by its unit inside the bracket.
在这种形式下，以 lg y 对 lg x 作图，斜率为 n，截距为 lg k，因此 k = 10截距。对数值无单位，因此坐标轴写作 "lg (T / s)" 这样的形式，把物理量除以单位放在括号内。
y = a·e−bx → ln y = ln a − bx
al form, plotting ln y against x gives a gradient of −b and an intercept of ln a, so a = eintercept gative gradient: quote the value with its sign and explain what the sign means in context.
对指数形式，以 ln y 对 x 作图，斜率为 −b，截距为 ln a，故 a = e截距。读负斜率时要小心：必须带上负号，并说明这个符号在物理情境中的含义。
Always check whether the question says "plot a graph of ... against ...". The first quantity goes on the y-axis and the second on the x-axis. Reversing them will make the gradient the reciprocal of the expected value and usually loses the analysis marks.
务必看清题目说"以……对……作图"。前一个量放纵轴，后一个量放横轴。若搞反了，斜率就会变成期望值的倒数，分析分通常也就丢了。
10. Evaluation and Improvements | 实验评估与改进建议
Evaluation marks come from judging how well your results support the relationship you were testing. A strong answer compares the percentage difference between your experimental value and the expected or theoretical value with the percentage uncertainty in your result.
评估分来自对"数据是否支持所检验关系"的判断。高质量答案会把实验值与理论（或期望）值的百分比差，与结果本身的百分比不确定度作比较。
If the percentage difference is smaller than the percentage uncertainty, the results support the relationship within experimental error. If it is much larger, there is likely to be a systematic error, and you should suggest what it could be and how to test for it.
若百分比差小于百分比不确定度，可以说结果在实验误差范围内支持该关系。若百分比差大得多，则很可能存在系统误差，你需要指出可能的来源以及如何检验它。
When discussing the scatter of points about the line of best fit, compare it with the size of the error bars. If the scatter is much larger than the error bars, the uncertainty in the readings has been underestimated; if it is much smaller, the uncertainty may have been overestimated.
讨论数据点相对最佳拟合线的离散程度时，要与误差棒的大小比较。若离散度远大于误差棒，说明读数不确定度被低估了；若远小于误差棒，则可能被高估了。
- State the problem clearly, for example "the trolley did not move at constant speed". | 清楚说明问题，例如"小车未保持匀速运动"。
- Give a specific improvement, such as "use a longer runway and a level track". | 给出具体改进，例如"使用更长的跑道并调平轨道"。
- Explain the effect, such as "this reduces the acceleration of the trolley and lowers the random error in the timing". | 说明影响，例如"这能减小小车的加速度，降低计时中的随机误差"。
- Suggest a check, for example "repeat with a different mass to see whether the discrepancy changes". | 提出验证方法，例如"换用不同质量重复实验，观察偏差是否随之改变"。
Vague statements such as "human error" or "the experiment was not accurate" earn nothing. Be specific, physical and quantitative: name the quantity, estimate its size, and link it to the result.
诸如"人为误差""实验不准确"这类空泛说法一分不得。要具体、物理化、定量化：说清是哪个量、估计有多大、如何影响结果。
11. Common Pitfalls and Marking Points | 常见失分点与评分要点
Most lost marks come from a small set of repeated mistakes rather than from a lack of physics knowledge. Knowing these in advance lets you avoid them under time pressure.
大多数丢分来自少数几个反复出现的错误，而不是物理知识不足。提前知道这些，就能在时间压力下主动避开。
- Units written inside every cell of a table instead of in the header. | 单位写在表格每个格子里，而不是写在表头。
- Inconsistent decimal places in a column of raw readings. | 同一列原始读数的小数位数不一致。
- Gradient taken from two data points instead of from the line of best fit. | 用两个数据点算斜率，而不是用最佳拟合直线。
- A triangle drawn so small that the gradient has a large uncertainty. | 三角形画得太小，导致斜率不确定度过大。
- Missing unit or wrong unit on the gradient and intercept. | 斜率和截距漏写单位或单位写错。
- Too many significant figures, such as quoting a gradient to six digits. | 有效数字过多，例如把斜率写到六位。
- Improvements given without a reason or without an effect on the result. | 只写改进，不给理由，也不说明对结果的影响。
- No comparison between the percentage difference and the percentage uncertainty. | 没有把百分比差与百分比不确定度作比较。
Examiners also reward clear presentation. Label your answers with the part letter, draw graphs in pencil with a ruler, and write the substitution before the answer, so that a small arithmetic slip does not cost you the method marks.
考官同样奖励清晰的呈现。标好小问字母，图像用铅笔和直尺绘制，先写代入式再写答案，这样即使计算有小失误，方法分也不会全丢。
12. Final Revision Checklist | 考前冲刺清单
In the last few days before the exam, focus on technique rather than new content. Practise reading a question, sketching the apparatus, and writing a table header correctly within five minutes.
考前几天应聚焦"技巧"而非新知识。练习在五分钟内读懂题目、画出装置简图、并正确写出表头。
- Can you state the absolute uncertainty for a metre rule, a vernier caliper and a stopwatch? | 你能说出米尺、游标卡尺和秒表的绝对不确定度吗？
- Can you combine percentage uncertainties for a formula involving powers? | 你能对含乘方的公式进行百分比不确定度合成吗？
- Can you linearise a relationship using lg or ln and identify the gradient and intercept? | 你能用 lg 或 ln 线性化关系，并指出斜率与截距的含义吗？
- Can you draw a good graph with correct axes, scale and line of best fit in under ten minutes? | 你能在十分钟内画出坐标轴、刻度、最佳拟合线都规范的图像吗？
- Can you write three specific improvements, each with a reason and an effect? | 你能写出三条具体改进，且每条都带理由和影响吗？
On the day, bring a sharp pencil, a clear ruler, an eraser, a calculator and a spare pen. Check the apparatus list given in the question, and use exactly the instruments the paper provides rather than assuming your school version.
考试当天带好削好的铅笔、透明直尺、橡皮、计算器和备用笔。核对题目给出的器材清单，严格使用试卷提供的仪器，不要想当然地按自己学校的配置作答。
Finally, remember that practical marks reward precision of language as much as precision of measurement. Say what you measured, how you measured it, what you found, how certain you are, and what could be improved. Do these five things consistently and your Paper 3 score will rise steadily.
最后要记住：实验题的分数既奖励"测量精度"，也奖励"语言精度"。说清你测了什么、怎么测的、得到了什么、有多确定、还能怎么改进。只要这五件事每次都做到，Paper 3 的分数就会稳步上升。
Published by TutorHao | A-Level Physics Revision Series | aleveler.com
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