📚 Year 8 CIE Statistics: Comparing UK University Entry Requirements | 英国大学申请要求对照
In Year 8 CIE Statistics you learn how to collect, organise, summarise and interpret data. One of the most motivating ways to apply those skills is to compare the entry requirements of for different UK universities. By turning published grade requirements into numbers, you can see at a glance which courses are more competitive and which are more accessible, and you can support your future choices with evidence instead of guesswork.
在 Year 8 CIE 统计学课程中，你将学习如何收集、整理、概括和解读数据。把这些技能应用到现实中最有动力的方式之一，就是比较不同英国大学的入学要求。把公开的成绩要求转换成数字之后，你一眼就能看出哪些专业竞争更激烈、哪些更容易达到，从而用证据而不是猜测来支持自己未来的选择。
ed example A*AA = 152, and the idea that some courses demand specific subject grades - and builds a complete Year 8 statistical toolkit around them: averages, spread, frequency tables, charts, sampling and common mistakes.
本文保留原有主题与全部原有信息点——A-level 成绩、UCAS 分数、从 A* = 56 到 E = 16 的分值表、A*AA = 152 的计算示例，以及"某些专业还会要求特定科目成绩"这一点——并围绕这些内容搭建一套完整的 Year 8 统计工具：平均数、离散程度、频数表、图表、抽样与常见错误。
1. Why Compare Entry Requirements? | 为什么要对比入学要求
University entry requirements are simply data. Each course publishes a minimum grade combination such as AAA or ABB, and each of those combinations can be converted into a single number. Once every course has a number, you can sort them, draw them, average them and compare them - which is exactly what statistical analysis means.
大学入学要求本质上就是数据。每个专业都会公布一个最低成绩组合，例如 AAA 或 ABB，而每一个组合都可以转换成一个数字。一旦每个专业都有了数字，你就可以排序、画图、求平均并加以对比——这正是统计分析的含义。
Comparing requirements also helps you set realistic targets. If your current predicted grades give 128 points and the courses you like sit between 136 and 168 points, statistics tells you the size of the gap instead of leaving you with a vague feeling of worry.
对比入学要求还能帮你设定现实的目标。如果你目前的预估成绩对应 128 分，而你喜欢的专业集中在 136 到 168 分之间，统计会明确告诉你差距有多大，而不是只留下一种模糊的焦虑感。
2. From A-level Grades to UCAS Points | 从 A-level 成绩到 UCAS 分数
UK universities usually express entry requirements either as A-level grades (for example AAA or A*AA) or as a UCAS tariff point total. The tariff converts each grade into a number, which makes calculation much easier.
英国大学通常用 A-level 成绩（例如 AAA 或 A*AA）或者 UCAS 总分来表达入学要求。UCAS 分数把每个等级转换成一个数字，计算起来就方便多了。
A* = 56 A = 48 B = 40 C = 32 D = 24 E = 16
20 points. Notice that the differences are not equal in size: the jump from A* to A is 8 points, and so is the jump from A to B, but the whole scale is squeezed towards the top, where competition is fiercest.
例如，一个要求 A*AA 的专业需要 56 + 48 + 48 = 152 分；而一个要求 BBB 的专业需要 40 + 40 + 40 = 120 分。注意这些差距并不完全等价：从 A* 到 A 是 8 分，从 A 到 B 也是 8 分，但整个量表越靠近顶端越拥挤，而那里恰恰是竞争最激烈的地方。
- A*A*A = 56 + 56 + 56 = 168
- A*AA = 56 + 48 + 48 = 152
- AAA = 48 + 48 + 48 = 144
- AAB = 48 + 48 + 40 = 136
- ABB = 48 + 40 + 40 = 128
- BBB = 40 + 40 + 40 = 120
- BBC = 40 + 40 + 32 = 112
Some courses also require specific grades in particular subjects, such as an A in Mathematics or a B in Chemistry. When comparing, you compare courses you may record either the overall grade combination or the total points; both methods are valid, but converting to points makes it far easier to calculate averages and draw graphs.
有些专业还要求特定科目达到特定成绩，例如数学要 A、化学要 B。对比时，你既可以记录整体的成绩组合，也可以记录总分；两种方法都成立，但换算成分数会更容易计算平均数和绘制图表。
3. Building a Clean Dataset | 建立一份干净的数据集
A dataset is only as good as its organisation. Choose six to ten courses, then record the same information for each one in the same order: university name, course name, published requirement, points total, and any required subject grade.
数据集的质量取决于它的组织方式。选择 6 到 10 个专业，然后按同样的顺序为每一个记录同样的信息：大学名称、专业名称、公布的要求、总分、以及是否有指定科目要求。
| University | Requirement | UCAS points |
| University A | A*A*A | 168 |
| University B | A*AA | 152 |
| University C | AAA | 144 |
| University D | AAB | 136 |
| University E | ABB | 120 |
| University F | BBB | 112 |
This table is a quantitative dataset because every entry is a number, and it is a small sample rather than a population: it contains six chosen courses, not all 30 000-plus courses in the UK. Good statisticians always state clearly what their sample is.
这张表是一个定量数据集，因为每一条记录都是数字；同时它只是一个小样本而不是总体：它包含被选中的 6 个专业，而不是英国全部三万多门专业。优秀的统计学习者总会明确说明自己的样本范围。
4. Measures of Central Tendency | 集中趋势的度量
rsities, the values 168, 152, 144, 136, 120 and 112 add to 832, and 832 ÷ 6 ≈ 138.7 points.
平均数、中位数和众数描述数据的中心。平均数是平均分配的结果：把所有数值相加，再除以个数。对上面六所大学来说，168、152、144、136、120、112 之和是 832，832 ÷ 6 ≈ 138.7 分。
Mean = Σx ÷ n = 832 ÷ 6 ≈ 138.7
lue that appears most often; here every value appears once, so there is no mode - a perfectly acceptable answer.
中位数是把数据排序后位于中间的值。有 6 个数值时，我们取第 3 和第 4 个的平均数：(144 + 136) ÷ 2 = 140 分。众数是出现次数最多的数值；这里每个数值都只出现一次，所以没有众数——这本身就是一个完全合理的答案。
- Mean is affected by every value, including very high ones.
- Median is resistant to extreme values, so it suits skewed data such as salaries or house prices.
- Mode is useful for categorical data, such as the most common required subject.
In our example the mean (138.7) is slightly below the median (140), which tells you the distribution is pulled gently downwards by the two lower values - a first hint of skew.
在这个例子中，平均数 138.7 略低于中位数 140，说明分布被两个较低的数值稍微向下拉了一点——这是偏斜的第一个信号。
5. Measures of Spread: Range and IQR | 离散程度的度量：极差与四分位距
Two courses can share the same mean yet be completely different in character. Spread measures tell you how far apart the values are, which is often more informative than the centre.
两个专业可能有相同的平均数，但性质完全不同。离散程度衡量数值之间相距多远，这往往比平均数本身提供更多信息。
Range = largest − smallest = 168 − 112 = 56
e ordered list into two halves: 112, 120, 136 and 144, 152, 168. Then Q₁ = 120, Q₃ = 152 and IQR = 152 − 120 = 32.
极差计算很快但很脆弱，因为一个不寻常的专业就能让它大幅变化。四分位距（IQR）可以避免这个问题。把排序后的数据分成两半：112、120、136 与 144、152、168。于是 Q₁ = 120，Q₃ = 152，IQR = 152 − 120 = 32。
48 = 200. All six values fall between 72 and 200, so there are no outliers in this sample.
检查异常值要用围栏：Q₁ − 1.5 × IQR = 120 − 48 = 72 与 Q₃ + 1.5 × IQR = 152 + 48 = 200。六个数值都落在 72 到 200 之间，所以这个样本中没有异常值。
6. Frequency Tables and Grouped Data | 频数表与分组数据
When the same requirement appears many times, a frequency table is much clearer than a long list. Suppose you survey ten courses in one subject and record their points totals with frequencies.
当相同的要求出现很多次时，频数表比一长串清单清晰得多。假设你调查某个学科的十个专业，并记录它们总分出现的频数。
| Points x | Frequency f | f × x |
| 112 | 2 | 224 |
| 120 | 3 | 360 |
| 136 | 4 | 544 |
| 144 | 1 | 144 |
cy, f = 4. The cumulative frequency tells you how many courses sit at or below each level.
估计平均数 = Σfx ÷ Σf = 1272 ÷ 10 = 127.2 分；众数要求是 136 分，因为该行频数最大，f = 4。累计频数则告诉你有多少个专业处于某一水平或以下。
For grouped data, such as "100-119 points, 120-139 points" ng saves space but loses detail, and grouped means are estimates rather than exact values.
对于分组数据，例如"100-119 分、120-139 分"，一定要用每组的组中值。120-139 的组中值是 (120 + 139) ÷ 2 = 129.5。分组可以节省篇幅，但会丢失细节，而且分组算出的平均数是估计值而非准确值。
7. Choosing the Right Chart | 选择合适的图表
A bar chart with gaps between bars is perfect for comparing separate universities, because the categories are not on a continuous scale. Put the universities on the horizontal axis and the UCAS points on the vertical axis, and start the axis at zero.
条形图（柱子之间有间隔）非常适合比较彼此独立的大学，因为这些类别并不落在连续尺度上。把大学放在横轴、UCAS 分数放在纵轴，并且纵轴一定要从 0 开始。
A box plot is better for comparing groups, because it shows the median, the quartiles, the IQR and any outliers at the same time. Drawing one box plot for "Russell Group style" courses and one for "modern university" courses immediately reveals which group has a higher median and which is more spread out.
箱线图更适合比较不同组别，因为它能同时显示中位数、四分位数、四分位距和异常值。为"一类高要求专业"和"一类要求较低的专业"各画一个箱线图，就能立刻看出哪一组的中位数更高、哪一组更分散。
- Discrete comparison of separate courses → bar chart
- Comparison of two or three groups → side-by-side box plots
- Relationship between two variables, such as required grade and offer rate → scatter graph
- Change over several years → line graph
Whatever chart you choose, label both axes, give the chart a title, include units, and state the sample size. A chart without a title and units is not a finished piece of statistical work.
无论选哪种图，都要给两个轴加标签、为图表加标题、标明单位，并说明样本量。没有标题和单位的图表不算完整的统计作品。
8. Making Fair Comparisons Between Courses | 如何在专业之间做出公平比较
Comparisons become unfair when you mix different kinds of data. A three-year course with an integrated foundation year, or a course with a portfolio interview, cannot be judged on UCAS points alone.
当你把不同性质的数据混在一起时，比较就会变得不公平。包含一年预科的三年制专业，或者需要作品集面试的专业，都不能只靠 UCAS 分数来评判。
Always compare like with like. Fix the subject, then compare universities; or fix the university, then compare subjects. Record the year of the data too, because requirements are reviewed annually and last year's figure may no longer apply.
一定要同类比同类。先固定学科再比较大学，或者先固定大学再比较学科。同时要记录数据的年份，因为入学要求每年都会被审核，去年的数字可能已经不再适用。
It also helps to report the difference from your own target. If your predicted total is 136 points, then a course at 152 points is a gap of 16 points - roughly one grade. Expressing a gap in grades makes it far more actionable than expressing it in raw points.
另外，报告与你自己目标的差距也很有帮助。如果你的预估总分是 136 分，那么要求 152 分的专业差距是 16 分——大致相当于一个等级。把差距换算成等级，比用原始分数表达要有用得多。
9. Sampling and Bias in Your Data | 抽样与数据偏差
Your dataset is a sample, and the way you choose it decides whether your conclusions are trustworthy. If you only look at the six most famous universities, your mean will be biased upwards and will not represent UK higher education as a whole.
你的数据集只是一个样本，而选择样本的方式决定了结论是否可信。如果你只看六所最著名的大学，算出的平均数就会偏高，不能代表英国高等教育的整体情况。
- Random sampling: pick courses using a random number generator from a full list - fair but sometimes impractical.
- Stratified sampling: split courses by subject area, then sample within each group - good for balanced coverage.
- Convenience sampling: use the courses you already know - quick but strongly biased.
Sample size matters too. For a mean to be reliable, a Year 8 project should aim for at least 20 to 30 items. A sample of 6 gives a rough estimate; a sample of 60 gives a much steadier one, because a single unusual course has far less influence.
样本量同样重要。要让平均数可靠，Year 8 的项目应尽量选取至少 20 到 30 个条目。6 个数据只能给出粗略估计；60 个数据则稳定得多，因为单个不寻常的专业影响力会小很多。
10. Common Mistakes and Misleading Graphs | 常见错误与误导性图表
Some errors appear again and again in school statistics projects. Knowing them in advance is the fastest way to raise your mark.
在学校统计项目中，有些错误反复出现。提前了解它们，是提高成绩最快的方式。
- Starting the vertical axis above zero, which exaggerates small differences between universities.
- Using a 3D pie chart, which distorts the apparent size of each sector.
- Confusing the mean with the median when the data are skewed.
- Reporting a mean without any measure of spread, so the reader cannot judge reliability.
- Mixing UCAS points with A-level grades in the same calculation.
- Forgetting to state the sample size, the year and the source of the data.
Extrapolation is another trap. If a course asked for 136 points three years in a row, it is tempting to predict 136 again next year, but admission requirements depend on applicant numbers, funding and policy - a straight-line forecast is an assumption, not a fact.
另一个陷阱是外推。如果一个专业连续三年都要求 136 分，人们很容易预测明年还是 136 分，但入学要求取决于申请人数、经费和政策——直线预测只是假设，不是事实。
11. Interpreting Your Results and Writing Conclusions | 解读结果与撰写结论
A strong conclusion answers the original question, quotes specific statistics and acknowledges limits. Compare this weak ending: "University A is better" with this strong one: "The six sampled courses have a mean of 138.7 UCAS points and a median of 140, with a range of 56 points. The highest requirement (University A, 168 points) is 56 points above the lowest (University F, 112 points), which is more than three grades of difference."
有力的结论会回答最初的问题、引用具体统计量，并承认局限。对比一下：薄弱结尾是"大学 A 更好"，而有力结尾是："所抽样的六个专业平均数为 138.7 个 UCAS 分数，中位数为 140，极差为 56 分。最高要求（大学 A，168 分）比最低要求（大学 F，112 分）高出 56 分，相当于三个等级以上的差距。"
Then add a sentence about reliability: "Because the sample contains only six courses chosen from one subject, the results may not generalise to all UK universities." This single sentence shows the examiner that you understand sampling, bias and the difference between a sample and a population.
然后再加上一句关于可靠性的说明："由于该样本仅包含来自一个学科方向的六个专业，结论未必能推广到所有英国大学。"这一句话就能让考官看出你理解抽样、偏差，以及样本与总体的区别。
Conclusion = answer + evidence + limitation
Finally, link the statistics back to yourself. If your predicted profile is 136 points, then most of the sampled courses (four out of six) are within one grade of your target, which is a realistic and reassuring finding built entirely from data.
最后，把统计结果与你自身联系起来。如果你的预估成绩是 136 分，那么样本中大多数专业（六个中的四个）都在你目标的一个等级之内，这是一个完全基于数据得出的、既现实又让人安心的结论。
12. Quick Revision Checklist | 快速复习清单
Use this checklist before you submit any comparison project. It covers every skill tested in Year 8 CIE Statistics and every data point from the original article.
在提交任何对比类项目之前，先过一遍这份清单。它覆盖 Year 8 CIE 统计学考查的全部技能，以及原文中的全部数据点。
- Can you convert a grade combination into UCAS points? A* = 56, A = 48, B = 40, C = 32, D = 24, E = 16.
- Can you show that A*AA = 56 + 48 + 48 = 152 and that A*A*A = 168?
- Can you calculate mean, median and mode, and explain when each is best?
- Can you calculate range, quartiles, IQR and outlier fences?
- Can you build a frequency table and estimate a mean from Σfx ÷ Σf?
- Can you choose between a bar chart and a box plot, and label both axes?
- Can you identify bias in a sample and suggest a fairer sampling method?
- Can you write a conclusion containing a statistic, an interpretation and a limitation?
Remember the original point of the exercise: some courses also require specific grades in particular subjects, such as an A in Mathematics. When that happens, the total point score is only part of the picture, and a good statistician reports both the total and the subject condition.
记住这项练习最初的意义：有些专业还会要求特定科目达到特定成绩，例如数学要 A。遇到这种情况，总分只是全貌的一部分，优秀的统计学习者会同时报告总分和科目条件。
Practise with real published requirements, keep your table tidy, state your sample, and your Year 8 statistics work will be both accurate and genuinely useful for your own future choices.
用真实公布的要求来练习，保持表格整洁，说明你的样本，你的 Year 8 统计作业就会既准确，又对自己未来的选择真正有帮助。
Published by TutorHao | Year 8 Statistics 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