📚 Interdisciplinary Mixed Questions Training for AQA Statistics | AQA 统计跨学科综合题型训练
In AQA GCSE Statistics you are rarely asked about statistics in the abstract. Papers are set in contexts borrowed from biology, geography, physics, economics, psychology and sports science, and marks are awarded for three things: choosing the right method, applying it accurately, and interpreting the result in that context. This revision guide expands our earlier training notes into a complete cross-curricular programme. It covers means, ranges and overlap, sampling in human geography, measurement uncertainty in physics, index numbers in economics, correlation in psychology, lines of best fit in sports science, the wording examiners reward, and it finishes with a full worked mixed question.
在 AQA GCSE 统计考试中，你很少会遇到纯粹的"统计"问题。试卷总是放在生物、地理、物理、经济、心理学和体育科学的真实情境里，得分点有三个：选对方法、算对结果、并把结果放回情境中解释。本复习指南把此前的训练笔记扩展成一套完整的跨课程训练方案：涵盖均值、极差与重叠判断，人文地理中的抽样，物理中的测量不确定度，经济中的指数，心理学中的相关性，体育科学中的最佳拟合线，考官认可的表述方式，最后以一道完整的综合题收尾。
1. The Statistical Enquiry Cycle in Cross-Curricular Settings | 跨课程情境中的统计探究循环
Every AQA GCSE Statistics paper is built around the statistical enquiry cycle: specify the problem, plan the investigation, collect the data, process and represent it, then interpret and discuss. Interdisciplinary contexts exist mainly to test the last two stages, because interpreting a result means saying what it means for the subject the data came from, not just for the numbers.
每份 AQA GCSE 统计试卷都围绕统计探究循环构建：明确问题、制订计划、收集数据、处理与呈现数据，然后解释与讨论。跨学科情境主要用来检验后两个环节，因为"解释结果"意味着要说明它对数据所来自的那个学科意味着什么，而不仅仅是对数字意味着什么。
The good news is that the statistical toolkit never changes. A question about seedling growth, rainfall in a river basin, repeated timing of a pendulum, the price of a basket of goods, memory scores after different amounts of sleep and 100 m sprint times can all require exactly the same sequence of skills: summarise, compare, quantify the relationship, then judge how reliable the conclusion is.
好消息是，统计工具箱本身从不改变。幼苗生长、流域降雨量、单摆计时的重复读数、一篮子商品的价格、不同睡眠时长后的记忆得分、100 米短跑成绩——这些问题可能要求完全相同的技能顺序：概括、比较、量化关系，然后判断结论有多可靠。
The table below shows the contexts that appear most often and the tools they are usually designed to test. Use it as a diagnostic checklist: if you cannot immediately say which tool a context is probing, that is the topic you need to revise next.
下表列出了最常出现的情境以及它们通常要考查的工具。把它当作诊断清单：如果某个情境出现时你不能立刻说出它在考哪个工具，那就是你下一步需要复习的主题。
| Subject context | 学科情境 | Typical data | 典型数据 | Tool most often required | 最常考的工具 |
|---|---|---|
| Biology | 生物 | Seedling height, pulse rate | 幼苗高度、脉搏 | Mean, range, IQR, comparing groups | 均值、极差、四分位距、组间比较 |
| Geography | 地理 | Rainfall, population density | 降雨量、人口密度 | Stratified sampling, comparative pie charts | 分层抽样、比较饼图 |
| Physics | 物理 | Repeated readings of length or time | 长度或时间的重复读数 | Mean, uncertainty, percentage error | 均值、不确定度、百分比误差 |
| Economics | 经济 | Prices over several years | 多年价格 | Index numbers, weighted index, percentage change | 指数、加权指数、百分比变化 |
| Psychology | 心理学 | Memory score vs hours of sleep | 记忆得分与睡眠时长 | Scatter graph, correlation, confounding variables | 散点图、相关性、混杂变量 |
| Sports science | 体育科学 | Training hours vs sprint time | 训练时长与冲刺成绩 | Line of best fit, interpolation, extrapolation | 最佳拟合线、内插、外推 |
2. Biology: Controlled Experiments and Data Collection | 生物：对照实验与数据收集
In biology, controlled experiments compare two or more groups. Statistical thinking helps you design the experiment, collect the data and compare the results. Always consider random allocation, sample size and control groups before you look at any numbers.
在生物学中，对照实验比较两个或多个组别。统计思维帮助你设计实验、收集数据并比较结果。在看任何数字之前，始终要考虑随机分配、样本量和对照组。
Random allocation means every subject or plant has an equal chance of being placed in any group, which prevents the researcher from unconsciously putting the strongest seedlings in the best light. A control group receives no treatment or a standard treatment, so you can separate the effect of the variable you are testing from everything else that is happening.
随机分配意味着每个受试者或植株被分到任何一组的机会相等，这可以防止研究者不自觉地把最壮的幼苗放进光照最好的一组。对照组不接受处理或只接受标准处理，这样你才能把所检验变量的效应与其他一切正在发生的事情区分开来。
A typical exam question presents data on the growth of seedlings under two different light conditions. You may need to calculate the mean and range for each group, then comment on whether the difference is meaningful. Remember to use the range as a simple measure of spread and to note any overlap between the two sets of values.
一个典型的考题会给出两种光照条件下幼苗生长的数据。你可能需要计算每组的均值和极差，然后说明差异是否有意义。记住用极差作为简单的离散程度度量，并注意两组数值之间是否有重叠。
Example: Group A (high light): 5.2 cm, 5.8 cm, 5.5 cm, 5.9 cm, 5.6 cm. Group B (low light): 4.1 cm, 4.5 cm, 4.0 cm, 4.3 cm, 4.6 cm. Calculate the mean of Group A.
示例：A 组（强光）：5.2 cm、5.8 cm、5.5 cm、5.9 cm、5.6 cm。B 组（弱光）：4.1 cm、4.5 cm、4.0 cm、4.3 cm、4.6 cm。计算 A 组均值。
Mean A = (5.2 + 5.8 + 5.5 + 5.9 + 5.6) ÷ 5 = 28.0 ÷ 5 = 5.6 cm
Mean B = (4.1 + 4.5 + 4.0 + 4.3 + 4.6) ÷ 5 = 21.5 ÷ 5 = 4.3 cm
the two ranges barely overlap: the smallest value in Group A is 5.2 cm, which is larger than the largest value in Group B at 4.6 cm. This suggests a real effect of light level, so in the exam you should link the statistical finding back to the biological context: more light produced taller seedlings, probably because light is needed for photosynthesis.
A 组极差为 5.9 − 5.2 = 0.7 cm，B 组为 4.6 − 4.0 = 0.6 cm。两组均值相差 5.6 − 4.3 = 1.3 cm，且两个极差几乎不重叠：A 组最小值 5.2 cm 已大于 B 组最大值 4.6 cm。这表明光照水平确实产生了影响，因此在考试中要把统计发现联系回生物情境：光照更充足使幼苗长得更高，很可能是因为光合作用需要光。
3. Judging Whether a Difference Is Meaningful | 判断差异是否有意义
A difference in means is not automatically a real difference. Examiners want a judgement based on the spread, the size of the sample and the context. Two rules cover most questions: compare the spread first, then compare the means in the light of that spread.
均值有差异并不自动等于存在真实差异。考官想要的是基于离散程度、样本量和情境作出的判断。两条规则可以覆盖大多数题目：先比较离散程度，再在该离散程度的背景下比较均值。
If the ranges or interquartile ranges overlap heavily, then a small difference in means could easily be produced by natural variation in the sample. If the ranges are short and completely separated, the difference is more convincing even with a modest sample size.
如果极差或四分位距重叠严重，那么均值的微小差异很可能是样本的自然变异造成的。如果极差很小且完全分离，即使样本量不大，差异也比较可信。
Sample size matters because larger samples reduce the influence of unusual individuals. A mean from 5 plants is fragile; a mean from 50 plants is much more stable. If a question tells you one group only has 4 data values, that is a prompt to mention reliability.
样本量很重要，因为样本越大，异常个体带来的影响越小。5 株植物算出的均值很脆弱，50 株植物算出的均值则稳定得多。如果题干告诉你某组只有 4 个数据，这其实是在提示你谈论可靠性。
Where your specification requires it, the standard deviation gives a more sensitive measure than the range because it uses every value rather than only the two extremes. A small standard deviation means the values cluster tightly around the mean, so a given difference in means becomes more meaningful.
在考纲要求的情况下，标准差比极差更灵敏，因为它使用了每一个数值，而不只是两个极端值。标准差小意味着数值紧密聚集在均值周围，此时给定的均值差异就更有意义。
- Step 1: calculate a measure of average (mean or median) for each group. | 第一步：为每组计算一个平均数（均值或中位数）。
- Step 2: calculate a measure of spread (range, IQR or standard deviation). | 第二步：计算一个离散程度度量（极差、四分位距或标准差）。
- Step 3: compare the averages, then use the spread to judge how reliable that comparison is. | 第三步：比较平均数，再用离散程度判断该比较有多可靠。
- Step 4: write one sentence that puts the conclusion back into the subject context. | 第四步：用一句话把结论放回学科情境中。
4. Geography: Sampling and Representing Spatial Data | 地理：抽样与空间数据表示
Geography questions are usually about how the data was obtained. A questionnaire handed out in one shopping centre, a count of cars taken at 9 a.m. on a Monday, or a survey of river pebbles at one point of the river all risk bias, and the exam expects you to name the bias and suggest an improvement.
地理类题目通常关注数据是如何取得的。只在一个购物中心发放问卷、只在周一上午 9 点统计车辆、只在河流某一点测量卵石，都会带来偏差，考试要求你指出偏差并给出改进建议。
Stratified sampling is the standard improvement when a population has clear subgroups, such as year groups in a school, districts in a city, or rock types along a river. The number taken from each stratum is proportional to its size, which makes the sample more representative.
当总体有明显子群时——例如学校里的年级、城市中的行政区、河流沿岸的岩石类型——分层抽样就是标准的改进方法。从每层抽取的数量与其规模成正比，样本因而更具代表性。
Sample from stratum = (stratum size ÷ population size) × sample size
s (1200 ÷ 6000) × 200 = 40 people, the second (1800 ÷ 6000) × 200 = 60, and the third (3000 ÷ 6000) × 200 = 100. Always check that your strata add up to the sample size: 40 + 60 + 100 = 200.
例如，某镇三个区的人口分别为 1200、1800 和 3000，你希望抽取 200 名居民作为样本。第一层贡献 (1200 ÷ 6000) × 200 = 40 人，第二层 (1800 ÷ 6000) × 200 = 60 人，第三层 (3000 ÷ 6000) × 200 = 100 人。一定要检查各层之和是否等于样本量：40 + 60 + 100 = 200。
When geographical data is shown as a comparative pie chart, the key rule is that it is the area of the pie, not the radius, that represents the frequency. So if one category is four times larger than another, its radius is only twice as long, because radius is proportional to the square root of the frequency.
当地理数据用比较饼图呈现时，关键规则是：代表频数的是扇形的面积，而不是半径。所以如果一个类别是另一个类别的四倍，它的半径只是两倍长，因为半径与频数的平方根成正比。
r₂ ÷ r₁ = √n₂ ÷ n₁
5. Physics: Measurement Error, Uncertainty and Repeated Readings | 物理：测量误差、不确定度与重复读数
Physics contexts test whether you understand that every measurement carries uncertainty. Repeating a reading and taking the mean reduces the effect of random error, which is unpredictable variation from one reading to the next. It cannot remove systematic error, which pushes all readings in the same direction, for example a ruler with a worn end.
物理情境考查的是你是否理解每次测量都带有不确定度。重复读数并取均值可以减小随机误差的影响，随机误差是读数之间不可预测的波动。但它无法消除系统误差，系统误差使所有读数朝同一方向偏移，例如一把磨损了起始端的尺子。
1.45 + 1.41 + 1.44 + 1.43) ÷ 5 = 7.15 ÷ 5 = 1.43 s, and the range is 1.45 − 1.41 = 0.04 s.
某单摆被计时五次，得到 1.42 s、1.45 s、1.41 s、1.44 s 和 1.43 s。均值为 (1.42 + 1.45 + 1.41 + 1.44 + 1.43) ÷ 5 = 7.15 ÷ 5 = 1.43 s，极差为 1.45 − 1.41 = 0.04 s。
A common convention is to quote the uncertainty as half the range, so the result is written as
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