📚 Drawing Conclusions and Evaluating Results in A-Level Physics | A-Level 物理：实验结论的得出与结果评估
In CIE A-Level Physics, Paper 3 (Practical) and Paper 5 (Planning and Analysis) require you not only to collect data but also to draw valid conclusions and evaluate the reliability of your results. This article provides a systematic framework for processing experimental data, deriving conclusions from graphs, and assessing uncertainties and errors in a way that satisfies the mark scheme.
在 CIE A-Level 物理考试中，Paper 3（实验卷）和 Paper 5（设计与分析卷）不仅要求你记录数据，还要求你得出有效结论并评估实验结果的可靠性。本文将提供一套系统的方法，帮助你处理实验数据、从图表中推导结论，并以符合评分标准的方式评估不确定性和误差。
1. Recording Raw Data and Creating Tables | 记录原始数据与制作表格
Raw data are the direct readings from instruments before any processing. They must be recorded in a table with correct headings, units, and an appropriate number of significant figures. For analogue instruments such as a ruler or protractor, record to one decimal place of the smallest division; for digital instruments, record all digits displayed.
原始数据是指未经任何处理的仪器直接读数。记录时必须使用规范的表格，包含正确的列标题、单位以及合理的有效数字位数。对于刻度尺、量角器等模拟仪器，估读到最小分度的下一位（通常为最小分度的十分之一）；对于数字式仪器，则记录显示屏上呈现的所有数字。
- Each column must have a quantity name and unit in the header, e.g. "l / cm".
- 每一列的标题必须包含物理量名称和单位，例如 "l / cm"（长度 / 厘米）。
- Repeat measurements at least twice for each independent variable value and calculate the average.
- 每组自变量至少重复测量两次并取平均值，以减小随机误差。
- Record raw data first, then processed data (averages, reciprocals, squares) in separate columns.
- 先记录原始数据，再将加工后的数据（平均值、倒数、平方等）放在后续独立列中。
- Do not round intermediate values too early; keep extra digits until the final answer.
- 不要过早舍入中间值；保留足够多的位数，直至最终结果才进行舍入。
2. Plotting Graphs and Drawing Lines of Best Fit | 绘制图表与最佳拟合线
Graphs serve two main purposes in A-Level practical work: they reveal the mathematical relationship between two variables and they average out random errors. When plotting, choose scales that use at least half of each axis. A graph that is too small introduces large percentage errors when reading gradients.
在 A-Level 实验中，图表有两个主要功能：揭示两个变量之间的数学关系，以及通过拟合过程平均掉随机误差。绘图时，坐标轴的比例应使数据点占据每个轴至少一半的长度。如果图表太小，读取斜率时将引入较大的百分比误差。
- Label each axis with the quantity name and unit, e.g. "T² / s²".
- 每个坐标轴都要标注物理量名称和单位，例如 "T² / s²"（周期的平方 / 秒²）。
- Use a sharp pencil and plot points as small crosses or dots with circles around them.
- 使用削尖的铅笔绘图，将数据点标示为小叉号或带圆圈的点。
- Draw a best-fit straight line using a transparent ruler, balancing points above and below the line.
- 用透明直尺画最佳拟合直线，使数据点均匀分布在直线两侧。
- Ignore anomalous points (outliers) only if you can identify a probable cause; otherwise include them in the fit.
- 只有在能够找出可能原因时才能忽略异常点（离群值）；否则应将其包含在拟合中。
Gradient = (y₂ − y₁) / (x₂ − x₁)
Choose two points on the best-fit line rather than two data points, and choose them far apart to reduce the percentage error in the gradient. The calculation triangle should be at least half the length of the drawn line.
应选择最佳拟合线上的两个点（而非两个数据点）来计算斜率，且两点间距应足够大，以减小斜率的百分比误差。计算三角形（两点间的水平和垂直距离）至少要达到所画直线长度的一半。
3. Linearising Non-Linear Relationships | 非线性关系的线性化
Many physical laws involve powers or exponentials. To test such a relationship graphically, you transform the data so that a straight-line graph can be plotted. Common transformations include y against x² , y against 1/x, or ln y against ln x.
许多物理定律涉及幂函数或指数关系。为了在实验数据中检验这类关系，我们需要对数据进行变换，使图形变为直线。常见的变换包括 y 对 x²、y 对 1/x、以及 ln y 对 ln x 等。
Consider a pendulum: the period T is related to length L by the equation:
以单摆为例：周期 T 与摆长 L 的关系满足方程：
T = 2π√L/g
e a straight line through the origin with gradient 4π²/g. From the gradient, g can be calculated directly.
两边平方得到 T² = (4π²/g) × L，因此 T² 对 L 的图象应为一条过原点的直线，斜率为 4π²/g。通过斜率即可直接计算出重力加速度 g。
- nship is y = a xⁿ, plot ln y against ln x; the slope gives n and the intercept gives ln a.
- 若关系式为 y = a xⁿ，则绘制 ln y 对 ln x 的图；斜率给出 n，截距给出 ln a。
- If the relationship is y = a eᵏˣ, plot ln y against x; the slope gives k and the intercept gives ln a.
- 若关系式为 y = a eᵏˣ，则绘制 ln y 对 x 的图；斜率给出 k，截距给出 ln a。
- Always state the transformed quantities in the table of processed data.
- 在数据处理表中，始终明确标注变换后的物理量（如 T²、ln y 等）。
4. Calculating Gradient and Intercept with Uncertainties | 计算斜率与截距及其不确定度
xis, or calculated by substituting the gradient and one point on the line into y = mx + c.
画出最佳拟合线后，利用线上两个相距较远的点计算斜率。y 截距则直接读取直线与 y 轴的交点，或者将斜率和线上任意一点的坐标代入直线方程 y = mx + c 中求解。
To estimate the uncertainty in the gradient, draw the steepest and least steep lines that still pass through most error bars. The uncertainty in the gradient is:
要估算斜率的不确定度，可以分别画出仍然穿过大多数误差条的最陡直线和最平缓直线。斜率的不确定度如下：
Δm = (mmax − mmin) / 2
Similarly, the uncertainty in the intercept is:
同样地，截距的不确定度为：
Δc = (cmax − cmin) / 2
| Quantity | 物理量 | Value | 数值 | Uncertainty | 不确定度 |
| Gradient m | 斜率 m | mbest | (mmax − mmin)/2 |
| Intercept c | 截距 c | cbest | (cmax − cmin)/2 |
When presenting final results, always quote the absolute uncertainty to one significant figure and round the measured value to match the decimal place of the uncertainty.
在呈现最终结果时，绝对不确定度通常保留一位有效数字，并将测量值舍入到与不确定度相同的小数位数。
5. Drawing Quantitative Conclusions from Graphs | 从图表得出定量结论
A conclusion must state whether the data support the hypothesized relationship and must include quantitative evidence. For example, if theory predicts T² ∝ L, you must check whether the graph is a straight line passing through the origin. If the intercept is not zero within its uncertainty, you should discuss this discrepancy.
结论必须说明数据是否支持预期的关系，并且必须包含定量证据。例如，如果理论预测 T² ∝ L，你需要检查图象是否为一条过原点的直线。如果截距在其不确定度范围内不为零，你应当讨论这一偏差的来源。
An acceptable conclusion has three components:
一个合格的结论包含三个要素：
- State the relationship found, e.g. "The graph of T² against L is a straight line passing through the origin, confirming T² ∝ L."
- 说明所发现的关系，例如："T² 对 L 的图象是一条过原点的直线，证实 T² ∝ L。"
- Quote the gradient with its uncertainty and compare it with the theoretical value.
- 给出斜率及其不确定度，并与理论值进行比较。
- 8 ± 0.2 m/s²) and discuss whether it agrees with the accepted value within the error bars.
- 给出最终计算出的物理常数（例如 g = 9.8 ± 0.2 m/s²），并讨论其是否在误差范围内与公认值一致。
Percentage difference = |measured value − accepted value| / accepted value × 100%
If the percentage difference is less than the percentage uncertainty of your measurement, the result is consistent with the accepted value.
如果百分比差异小于测量的百分比不确定度，则说明实验结果与公认值在误差范围内一致。
6. Comparing Theories and Experimental Values | 比较理论与实验值
Quantitative comparison is a key skill in Paper 5. You must decide whether a measured value agrees with a theoretical prediction or with a known accepted value. The comparison must be made using uncertainties, not simply by looking at whether the numbers are close.
定量比较是 Paper 5 中的关键技能。你需要判断测量值是否与理论预测或已知的公认值一致。比较必须基于不确定度来进行，而不能仅仅观察数值是否接近。
ange of your measurement is from 9.3 to 9.9 m/s². Since 9.81 lies within this range, your result is consistent with the accepted value.
例如，若 g 的理论值为 9.81 m/s²，而你的实验结果为 g = 9.6 ± 0.3 m/s²，则你的测量范围是 9.3 到 9.9 m/s²。由于 9.81 落在这个范围内，你的结果与公认值是一致的。
- e accepted value falls within (measured value ± uncertainty), conclude there is agreement.
- 如果公认值落在（测量值 ± 不确定度）范围内，则结论为两者一致。
- If the accepted value lies outside this range, state that there is a systematic discrepancy and suggest causes.
- 如果公认值落在这个范围之外，则说明存在系统性差异，并应提出可能的原因。
- Never conclude "the results are correct" — state only that the results are consistent or inconsistent with theory.
- 永远不要下结论说"结果是正确的"——只能说明结果与理论一致或不一致。
7. Evaluating Uncertainties and Limitations | 评估不确定度与实验局限
Evaluation is the process of identifying the major sources of error in an experiment and estimating their effects on the final result. Random errors cause scatter in the data and can be reduced by repeating measurements and using more precise instruments. Systematic errors cause all readings to shift in one direction and are not reduced by repetition.
评估是指识别实验中主要误差来源并估计它们对最终结果影响的过程。随机误差导致数据出现散布，可以通过重复测量和使用更精密的仪器来减小。系统误差使所有读数朝同一方向偏移，重复测量并不能减小这类误差。
| Type | 类型 | Source | 来源 | Reduction | 减小方法 |
| Random | 随机 | Judgement in reading scale, timing reaction, parallax error | Repeat and average, use a vernier/digital instrument, view scale perpendicularly |
| Systematic | 系统 | Zero error, instrument calibration, heat loss, friction | Calibrate the instrument, correct for zero error, improve insulation, use smoother surfaces |
For each experimental limitation, you should describe it, state how it affects the results (direction and magnitude if possible), and suggest a practical improvement.
对于每一项实验局限，你应当描述其具体内容，说明它如何影响结果（尽可能指出影响方向和大小），并提出切实可行的改进方案。
8. Identifying Systematic Errors | 识别系统性误差
Systematic errors are often more serious than random errors because they cannot be averaged out. Common systematic errors in A-Level experiments include: zero error on a balance or stopwatch, heat loss in calorimetry experiments, air resistance on falling objects, and friction in dynamics experiments.
系统误差通常比随机误差更严重，因为它们无法通过平均来消除。A-Level 实验中常见的系统误差包括：天平的零点误差、秒表的零点误差、量热实验中的热量散失、落体运动中的空气阻力，以及动力学实验中的摩擦力。
To identify a systematic error, compare the plotted line with the theoretical expectation. If the graph of y against x is straight but does not pass through the origin when theory predicts it should, there may be a systematic offset. For example, if a spring does not obey Hooke's law from the start due to a pre-load, the intercept of F against x will be non-zero.
要识别系统误差，可以将所绘制的图线与理论预期进行比较。如果 y 对 x 的图象为直线，但理论预测应过原点时它却不过原点，则可能存在系统性偏移。例如，若弹簧因预加载而从一开始就不满足胡克定律，F 对 x 的图截距将不为零。
9. Suggesting Improvements to the Experiment | 提出实验改进方案
Improvements must be specific to the apparatus and procedure described in your experiment. Generic statements like "use more accurate equipment" will not earn full marks. Instead, name the exact piece of equipment and explain how it reduces the specific error.
改进方案必须针对你实验中所使用的仪器和操作步骤，提出具体的建议。像"使用更精密的仪器"这类泛泛之谈无法获得满分。相反，你应该指明具体的仪器名称，并解释它如何减少特定的误差。
- Replace a metre ruler with a vernier caliper or micrometer screw gauge to measure small lengths more precisely.
- 用游标卡尺或螺旋测微器取代米尺，以更精确地测量较小的长度。
- Use a light gate and digital timer instead of a stopwatch to eliminate human reaction time.
- 使用光电门和数字计时器代替秒表，以消除人的反应时间误差。
- Use a data logger to record temperature continuously, minimising heat loss between readings.
- 使用数据记录器连续记录温度，减少两次读数之间的热量散失。
- For pendulum experiments, release the bob from a clamped electromagnet to ensure consistent starting conditions.
- 对于单摆实验，使用夹持的电磁铁释放摆球，以确保每轮实验的初始条件一致。
- Repeat readings at each value and calculate the mean to reduce random uncertainty.
- 在每个数据点处重复测量并计算平均值，以减小随机不确定度。
10. Writing an Effective Conclusion and Evaluation Paragraph | 写出有效的结论与评估段落
The mark scheme for CIE A-Level practicals often awards separate marks for conclusion, uncertainty analysis, and improvements. A high-scoring evaluation paragraph should be structured as follows: state the result, quote the uncertainty, compare with the expected value, identify the largest source of error, and suggest an improvement that targets that source specifically.
CIE A-Level 实验题的评分标准通常会分别给结论、不确定度分析和改进建议打分。一个高分的评估段落应按照以下结构展开：说明实验结果，给出不确定度，与预期值比较，指出最大的误差来源，并针对该来源提出具体的改进方案。
An example of a well-written conclusion:
一个写得好的结论示例：
" ately ±0.2 s when starting and stopping the stopwatch, which gives an uncertainty of about 2.5% ld be improved by using a light gate timing system triggered by the bob crossing a sensor."
"最佳拟合线的斜率为 4.05 ± 0.12 m/s²，由此得到 g = 9.72 ± 0.29 m/s²。公认值 9.81 m/s² 位于不确定度范围内，因此实验结果与重力加速度的理论值一致。最大的误差来源是按动秒表开始和结束时约 ±0.2 s 的反应时间，这给周期带来了约 2.5% 的不确定度。使用由摆球通过传感器触发的光电门计时系统可以改进这一点。"
Published by TutorHao | 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