Year 13 OCR Statistics: A Statistical Analysis of UK University Entry Requirements | Year 13 OCR 统计学：英国大学申请要求的统计分析

📚 Year 13 OCR Statistics: A Statistical Analysis of UK University Entry Requirements | Year 13 OCR 统计学：英国大学申请要求的统计分析

As Year 13 students following the OCR A Level Mathematics specification with Statistics, you are developing the ability to collect, analyse and interpret data – skills that are invaluable not only for your exams but also for making informed decisions about your future. One of the most pressing decisions is where to apply for university. By treating university entry requirements as a dataset, we can use the statistical techniques from the OCR syllabus to compare offers, quantify variability and even test hypotheses about the competitiveness of different institutions. This article guides you through a full statistical investigation into UK entry requirements for Statistics and related degrees, blending revision of core topics with practical UCAS insights.

作为学习OCR A Level数学（含统计学）课程的Year 13学生，你们正在培养收集、分析和解释数据的能力——这些技能不仅对考试至关重要，也有助于你们就未来做出明智的决定。其中最紧迫的决定之一便是申请哪所大学。通过将大学入学要求视为一个数据集，我们可以运用OCR考纲中的统计方法来比较录取条件、量化差异性，甚至检验关于不同院校竞争力的假设。本文将引导你完成一项针对英国统计及相关专业入学要求的完整统计探究，既巩固核心知识点，又提供实用的UCAS洞察。

1. Data Collection: Entry Requirements for Statistics Degrees | 数据收集：统计类学位的入学要求

Any statistical investigation begins with data collection. We need to identify the population – here, all UK universities offering BSc Statistics or closely related programmes such as Mathematics with Statistics or Data Science – and then draw a representative sample. For our analysis, we selected nine institutions, deliberately including both Russell Group and non‑Russell Group members to explore contrasts. The typical offer (expressed in A Level grades) was recorded from official university websites in October 2025. To enable numerical analysis, each grade combination was converted into tariff points using a simplified equivalence: A* = 56, A = 48, B = 40, C = 32. This produces a single continuous‑type discrete variable that captures the overall demand.

任何统计探究都始于数据收集。我们需要确定总体——这里指所有提供统计学学士（BSc Statistics）或相关专业（如数学与统计、数据科学）的英国大学——然后抽取一个具有代表性的样本。在本次分析中，我们选择了九所院校，特意包含了罗素集团和非罗素集团成员，以便探究它们之间的差异。典型录取条件（以A Level成绩表示）于2025年10月从各校官网获取。为了进行数值分析，每组成绩组合按照简化等价转换为tariff points：A* = 56, A = 48, B = 40, C = 32。由此产生一个连续的离散型变量，用以刻画总体要求水平。

University	Russell Group	Typical Offer	Tariff Points (x)
Imperial College London	Yes	AAA	160
University of Oxford	Yes	A*AA	152
University of Warwick	Yes	A*AA	152
University of Southampton	Yes	AAA	144
University of Leeds	Yes	AAB	136
University of Leicester	No	AAB	136
University of Kent	No	ABB	128
University of Greenwich	No	BBB	120
University of Hertfordshire	No	BBC	112

This table constitutes our raw data. Notice that the variable “Tariff Points” is discrete and measured on a ratio scale: a score of 160 is genuinely twice as demanding as 80 in the underlying tariff framework, and zero would mean no points, although in practice we never observe values that low for degree entry.

上表构成了我们的原始数据。请注意，“tariff points”这一变量是离散的，并且属于比率尺度：在基础计分框架下，160分的难度确实两倍于80分，而零分意味着没有积分，尽管在实践中我们不会在学位入学要求中见到如此低的数值。

2. Types of Variables and Data Visualisation Considerations | 变量类型与数据可视化考量

The dataset contains two variables: tariff points (quantitative, discrete) and group membership (categorical: Russell Group or non‑Russell Group). Before summarising, we must consider the appropriate graphs. For a single quantitative variable, a box plot gives a clear picture of the five‑number summary and highlights outliers. When comparing two groups, side‑by‑side box plots are extremely effective. A histogram would also be useful for showing the shape of the tariff distribution if our sample were larger, but with only nine observations the box plot remains the better choice.

该数据集包含两个变量：tariff points（定量、离散）和组别归属（分类变量：罗素集团或非罗素集团）。在开始汇总之前，我们需要考虑合适的图形。对于单个定量变量，箱线图能清晰展示五数概括，并突出显示异常值。当比较两个组别时，并排箱线图非常有效。如果样本量较大，直方图也有助于显示tariff分数的分布形态，但在仅有九个观测值的情况下，箱线图是更优的选择。

OCR Statistics expects you to understand the distinction between explanatory and response variables. In our investigation, the university group (Russell Group status) can be treated as the explanatory variable, and the tariff points as the response. This naturally leads to comparative summary statistics.

OCR统计学要求你能够区分解释变量与响应变量。在我们的探究中，大学组别（是否为罗素集团）可被视为解释变量，而tariff points则为响应变量。这自然引出了比较性的汇总统计。

3. Measures of Central Tendency: Average Entry Tariff | 集中趋势的度量：平均入学积分

We start by calculating the sample mean tariff for the whole dataset and then for each subgroup. Let x denote tariff points. For all nine universities:

我们从计算整个数据集以及各子组的样本平均积分开始。设 x 表示tariff points。对于全部九所大学：

x̄ = (160 + 152 + 152 + 144 + 136 + 136 + 128 + 120 + 112) / 9 = 137.78 (to 2 d.p.)

Now for the five Russell Group universities (RG):

对于五所罗素集团大学（RG）：

x̄ₑₔ = (160 + 152 + 152 + 144 + 136) / 5 = 148.8

For the four non‑Russell Group universities (NRG):

对于四所非罗素集团大学（NRG）：

x̄ₙₙₔ = (136 + 128 + 120 + 112) / 4 = 124.0

The median is also worth reporting because the mean can be influenced by extreme values, although our small symmetric‑ish sample is fairly well‑behaved. Ordered tariff values: 112, 120, 128, 136, 136, 144, 152, 152, 160. The median (5th value) is 136. For the RG group, median = 152; for the NRG group, median = 124. These measures already reveal a noticeable difference, with the Russell Group offers typically 24–25 tariff points higher.

中位数同样值得报告，因为均值可能会受极端值的影响，不过我们这个小而大致对称的样本表现得较为规整。排序后的tariff值为：112, 120, 128, 136, 136, 144, 152, 152, 160。其中位数（第5个值）为136。对于RG组，中位数为152；对于NRG组，中位数为124。这些指标已经揭示出显著的差异：罗素集团的录取条件普遍高出24–25个tariff points。

4. Measures of Dispersion: Range, IQR and Standard Deviation | 离散程度的度量：极差、四分位距与标准差

Centre alone cannot tell the full story; we need measures of spread to understand how consistent the requirements are within each group. The range for all nine universities is 160 − 112 = 48. However, the range is sensitive to extremes. We therefore calculate the interquartile range (IQR). Q₁ (25th percentile) is the average of the 2nd and 3rd ordered values: (120 + 128)/2 = 124. Q₃ (75th percentile) is the average of the 7th and 8th values: (152 + 152)/2 = 152. Hence IQR = 152 − 124 = 28. For the RG group: ordered 136,144,152,152,160 → Q₁ = (136+144)/2 = 140, Q₃ = (152+152)/2 = 152, IQR = 12. For the NRG group: 112,120,128,136 → Q₁ = 116, Q₃ = 132, IQR = 16. The Russell Group requirements are more tightly clustered than the non‑Russell Group offers, which show greater relative variability.

仅有集中趋势无法反映全貌；我们还需要离散程度的度量，以理解各组内部录取要求的一致程度。全部九所大学的极差为 160 − 112 = 48。然而，极差对极端值敏感。因此我们计算四分位距（IQR）。Q₁（第25百分位数）为第2、3个排序值的平均：(120 + 128)/2 = 124。Q₃（第75百分位数）为第7、8个值的平均：(152 + 152)/2 = 152。因此IQR = 152 − 124 = 28。对于RG组：排序为136,144,152,152,160 → Q₁ = (136+144)/2 = 140, Q₃ = 152, IQR = 12。对于NRG组：112,120,128,136 → Q₁ = 116, Q₃ = 132, IQR = 16。罗素集团的要求比非罗素集团更为集中，后者展现出较大的相对变异性。

The sample standard deviation s is the square root of the variance. We use the formula:

样本标准差 s 是方差的平方根。我们使用如下公式：

s = √[ Σ(xᵢ − x̄)² / (n − 1) ]

For the whole sample, x̄ ≈ 137.78. The squared deviations are: (160−137.78)² = 490.0; (152−137.78)² = 203.0 (twice); (144−137.78)² = 38.7; (136−137.78)² = 3.2 (twice); (128−137.78)² = 95.6; (120−137.78)² = 316.0; (112−137.78)² = 664.6. Summing gives ≈ 2017.3. Dividing by 8 (n−1) yields variance ≈ 252.16, so s ≈ √252.16 ≈ 15.9. The RG group alone: mean 148.8, squared deviations sum = (160−148.8)² + 2×(152−148.8)² + (144−148.8)² + (136−148.8)² = 125.44 + 2×10.24 + 23.04 + 163.84 = 332.8. Dividing by 4 gives variance 83.2, s ≈ 9.1. The NRG group: mean 124, deviations sum = (136−124)² + (128−124)² + (120−124)² + (112−124)² = 144+16+16+144 = 320. Divide by 3 gives variance ≈ 106.7, s ≈ 10.3. Russell Group offers have a smaller standard deviation, reinforcing the IQR finding.

全样本：均值137.78，偏差平方和 ≈ 2017.3，除以8得方差252.16，s ≈ 15.9。RG组：均值148.8，偏差平方和 = 332.8，除以4得方差83.2，s ≈ 9.1。NRG组：均值124，偏差平方和 = 320，除以3得方差106.7，s ≈ 10.3。罗素集团录取条件的标准差更小，进一步印证了IQR的发现。

5. Box Plots and Comparing Distributions | 箱线图与分布比较

Based on the five‑number summaries we can construct box plots. For all data: min=112, Q₁=124, median=136, Q₃=152, max=160. For RG: min=136, Q₁=140, median=152, Q₃=152, max=160. For NRG: min=112, Q₁=116, median=124, Q₃=132, max=136. Overlaying these distributions makes the contrast immediate: the RG box is shifted to higher tariff scores and is more compact; the NRG box shows a lower location and a slightly longer whisker, but no outliers in either group. This visual evidence strongly suggests that the two groups come from different underlying populations, a claim that can be formally assessed with a hypothesis test.

基于上述五数概括，我们可以绘制箱线图。全部数据：min=112, Q₁=124, median=136, Q₃=152, max=160。RG组：min=136, Q₁=140, median=152, Q₃=152, max=160。NRG组：min=112, Q₁=116, median=124, Q₃=132, max=136。将这些分布并排叠加后，差异一目了然：RG箱体整体向右移动且更加紧凑；NRG箱体位置偏低，触须稍长，但两组均无异常值。这一直观证据强烈表明，这两组数据来自不同的总体，该论断可通过假设检验予以正式评估。

6. Probability Distributions: Modelling Individual Admission Chances | 概率分布：个人录取机会建模

In OCR Statistics, you learn that the normal distribution can model continuous variables, and the binomial distribution models the number of successes in a fixed number of trials. While tariff points are discrete, their sum over three subjects is often treated as approximately normal if the sample were larger. However, here we can use binomial probability to consider an individual student’s chance of meeting a typical offer. Suppose a student with predicted A*AA applies to a Russell Group university whose

Published by TutorHao | Year 13 统计 Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）

Year 13 OCR Statistics: A Statistical Analysis of UK University Entry Requirements | Year 13 OCR 统计学：英国大学申请要求的统计分析