📚 Regionalism and the EU: A Statistical Modelling Approach | 区域主义与欧盟：统计建模方法
In A-Level Mathematics, statistical modelling gives us powerful tools to investigate real-world questions such as how regionalism within the European Union can be measured, compared and tested. This article treats 'regionalism and the EU' as a data-driven problem: we use Edexcel-style statistical techniques to examine regional economic convergence, trade integration, funding allocation and policy preferences. The focus is not on political theory but on the mathematical methods that allow us to draw reliable conclusions from numerical evidence.
在 A-Level 数学中，统计建模为我们提供了强大的工具来研究现实问题，例如如何测量、比较和检验欧盟内部的区域主义。本文将“区域主义与欧盟”视为一个数据驱动的问题：我们使用 Edexcel 风格的统计技术来考察区域经济趋同、贸易一体化、资金分配和政策偏好。重点不是政治理论，而是使我们能够从数值证据中得出可靠结论的数学方法。
1. Defining Regionalism in Measurable Terms | 用可测量的方式定义区域主义
To apply statistical methods, we first need to transform the broad concept of regionalism into quantitative indicators. Common measurable variables include regional GDP per capita, intra-EU trade share as a percentage of total trade, net EU funding received per capita, and regional unemployment rates. A discrete variable such as the number of EU-funded projects in a region can be modelled using count distributions, while continuous variables such as GDP growth rate can be summarised using means and standard deviations.
为了应用统计方法，我们首先需要将宽泛的区域主义概念转化为定量指标。常见的可测量变量包括区域人均 GDP、欧盟内部贸易占贸易总额的百分比、人均获得的欧盟净资金以及区域失业率。诸如某地区欧盟资助项目数量这样的离散变量可以用计数分布建模，而像 GDP 增长率这样的连续变量则可以用均值和标准差来概括。
- Continuous variable example: regional GDP per capita in euros, measured to the nearest €100.
- 连续变量示例：区域人均 GDP（欧元），精确到 100 欧元。
- Discrete variable example: number of cross-border cooperation projects in a region per year.
- 离散变量示例：某地区每年跨境合作项目的数量。
2. Data Collection and Sampling Methods | 数据收集与抽样方法
Reliable statistical inference depends on how the data is collected. In studying EU regionalism, a census of all NUTS 2 regions is often possible because there are about 240 such regions. However, if we only sample a subset, we must decide between simple random sampling, stratified sampling by member state, or cluster sampling by broader regions. Stratified sampling is particularly useful because it ensures that smaller countries are represented in proportion to their number of regions, reducing bias when comparing regional policies.
可靠的统计推断取决于数据的收集方式。在研究欧盟区域主义时，对全部 NUTS 2 区域进行普查通常是可行的，因为大约有 240 个这样的区域。然而，如果我们只抽取一个子集，就必须在简单随机抽样、按成员国分层抽样或按更大区域整群抽样之间做出选择。分层抽样特别有用，因为它能确保较小的国家按其区域数量比例被代表，从而在比较区域政策时减少偏差。
Sampling bias can arise if only regions that receive large EU grants are selected for analysis. To avoid this, the sampling frame should include all eligible regions, and each region should have a known non-zero probability of selection.
如果只选择获得大额欧盟拨款的区域进行分析，就会产生抽样偏差。为了避免这种情况，抽样框应包括所有符合条件的区域，并且每个区域都应有已知的非零被选概率。
3. Measures of Central Tendency and Dispersion | 集中趋势与离散程度的度量
Once regional data is collected, we summarise it using the mean, median and mode. The mean regional GDP per capita gives an overall average, but it is sensitive to extreme values such as Luxembourg or inner London. The median is often preferred for skewed income data because it is not affected by outliers. The interquartile range (IQR) and standard deviation measure how spread out the regional values are.
一旦收集了区域数据，我们就使用均值、中位数和众数来概括它们。区域人均 GDP 的均值给出了总体平均水平，但它对卢森堡或伦敦内城等极端值很敏感。对于偏斜的收入数据，中位数通常更受青睐，因为它不受异常值影响。四分位距（IQR）和标准差衡量区域值的分散程度。
Mean = (Σxᵢ) / n , Standard deviation s = √[ Σ(xᵢ − x̄)² / (n − 1) ]
For EU regional data, a large standard deviation indicates strong regional inequality, while a small standard deviation suggests convergence towards similar economic performance.
对于欧盟区域数据，较大的标准差表明区域不平等程度高，而较小的标准差则表明经济表现趋于相似，即存在趋同。
4. Probability Distributions for Regional Indicators | 区域指标的概率分布
Many regional indicators can be modelled using standard probability distributions. The number of EU-funded projects in a region over a fixed period can be modelled by a Poisson distribution if events occur independently at a constant average rate. For example, if a region receives an average of 3.2 projects per year, we can calculate the probability of receiving exactly 5 projects.
许多区域指标可以用标准概率分布来建模。如果事件独立发生且平均发生率恒定，则某地区在固定时期内欧盟资助项目的数量可以用泊松分布建模。例如，如果一个地区平均每年获得 3.2 个项目，我们可以计算恰好获得 5 个项目的概率。
X ~ Po(λ), P(X = r) = (e⁻λ λʳ) / r!
With λ = 3.2 and r = 5, P(X = 5) = (e⁻³·² × 3.2⁵) / 5! ≈ 0.114. This means there is about an 11.4% chance of exactly five projects in that year.
当 λ = 3.2 且 r = 5 时，P(X = 5) = (e⁻³·² × 3.2⁵) / 5! ≈ 0.114。这意味着该年恰好有五个项目的概率约为 11.4%。
5. Correlation Between Trade Integration and GDP Growth | 贸易一体化与 GDP 增长之间的相关性
Regionalism in the EU is often associated with increased intra-EU trade. To test this, we can calculate the product moment correlation coefficient (PMCC) between a region's intra-EU trade share and its annual GDP growth rate. The PMCC, denoted by r, ranges from −1 to +1. A positive r close to +1 would suggest that regions more integrated into EU trade tend to grow faster.
欧盟的区域主义通常与欧盟内部贸易的增加有关。为了检验这一点，我们可以计算一个地区的欧盟内部贸易份额与其年 GDP 增长率之间的积矩相关系数（PMCC）。PMCC 用 r 表示，取值范围为 −1 到 +1。正的 r 接近 +1 表明欧盟贸易一体化程度更高的地区往往增长更快。
r = Sxy / √Sxx × Syy , where Sxy = Σ(xᵢ − x̄)(yᵢ − ȳ)
Remember that correlation does not imply causation. A high correlation may be due to a third factor such as infrastructure quality or education levels. Nevertheless, correlation analysis is a useful first step in statistical modelling of regionalism.
请记住，相关性并不意味着因果关系。高相关性可能是由于基础设施质量或教育水平等第三个因素造成的。尽管如此，相关性分析是区域主义统计建模中很有用的第一步。
6. Linear Regression Modelling of Regional Convergence | 区域趋同的线性回归建模
If a scatter diagram shows a roughly linear relationship, we can fit a least squares regression line. For example, we might regress regional GDP growth (y) on initial GDP per capita (x). A negative slope would support the convergence hypothesis: regions with lower initial GDP tend to grow faster, reducing regional disparities over time.
如果散点图显示大致呈线性关系，我们可以拟合一条最小二乘回归线。例如，我们可以将区域 GDP 增长（y）对初始人均 GDP（x）进行回归。负斜率将支持趋同假说：初始 GDP 较低的地区往往增长更快，从而随时间缩小区域差距。
y = a + bx , b = Sxy / Sxx , a = ȳ − bx̄
the estimated regression equation is y = 3.8 − 0.35x. This means that for every additional € percentage points. Such a result is commonly found in studies of EU regional convergence.
假设估计的回归方程为 y = 3.8 − 0.35x。这意味着初始人均 GDP 每增加 1,000 欧元，预测的年增长率就下降 0.35 个百分点。这一结果在欧盟区域趋同研究中很常见。
| Initial GDP per capita (€1,000), x | Predicted growth rate (%), y |
|---|---|
| 10 | 3.8 − 3.5 = 0.3 |
| 20 | 3.8 − 7.0 = −3.2 |
| 30 | 3.8 − 10.5 = −6.7 |
The negative predictions for high-income regions indicate that the model is only valid within the observed range of data; extrapolation beyond that range is unreliable.
高收入区域的负预测表明该模型仅在观测数据范围内有效；超出该范围的外推不可靠。
7. Hypothesis Testing for Regional Convergence | 区域趋同的假设检验
and the alternative hypothesis is H₁: b < 0 (regions with lower initial GDP grow faster).
我们可以检验趋同回归中负斜率是否具有统计显著性。可以对回归系数 b 使用单尾 t 检验。原假设为 H₀: b = 0（无趋同），备择假设为 H₁: b < 0（初始 GDP 较低的区域增长更快）。
Test statistic t = (b − 0) / SE(b) , where SE(b) is the standard error of b
If the calculated t-value is less than the critical value from the t-distribution with n − 2 degrees of freedom at the 5% significance level, we reject H₀ and conclude that there is evidence of regional convergence. Otherwise, we do not have sufficient evidence.
如果在 5% 显著性水平下，计算出的 t 值小于自由度为 n − 2 的 t 分布的临界值，我们就拒绝 H₀，并得出结论：有区域趋同的证据。否则，我们没有足够的证据。
8. Normal Distribution and Confidence Intervals | 正态分布与置信区间
Many regional indicators, such as the percentage of intra-EU trade, can be approximated by a normal distribution for large samples. We can construct a confidence interval for the mean intra-EU trade share using the formula based on the normal distribution when the population variance is known, or using the t-distribution when it is estimated from the sample.
许多区域指标，例如欧盟内部贸易的百分比，在大样本情况下可以近似用正态分布来描述。当总体方差已知时，我们可以使用基于正态分布的公式构建平均欧盟内部贸易份额的置信区间；当方差从样本中估计时，则使用 t 分布。
95% CI for μ: x̄ ± z × σ / √n , where z = 1.96 for a two-tailed 95% interval
For instance, if a sample of 50 regions has a mean intra-EU trade share of 62% and a known population standard deviation of 8%, the 95% confidence interval is 62 ± 1.96 × 8 / √50, which gives 62 ± 2.22, i.e. [59.78%, 64.22%].
例如，如果 50 个区域的样本平均欧盟内部贸易份额为 62%，已知总体标准差为 8%，那么 95% 置信区间为 62 ± 1.96 × 8 / √50，即 62 ± 2.22，结果为 [59.78%, 64.22%]。
9. Time Series Analysis of Regional Integration | 区域一体化的时间序列分析
Regionalism develops over time, so we often study time series data such as annual intra-EU trade volume or regional GDP. A time series can be decomposed into trend, seasonal variation and random fluctuations. We can use moving averages to smooth out short-term fluctuations and identify the underlying trend in regional integration.
区域主义随时间发展，因此我们经常研究时间序列数据，例如年度欧盟内部贸易额或区域 GDP。时间序列可以分解为趋势、季节变动和随机波动。我们可以使用移动平均来平滑短期波动，并识别区域一体化的潜在趋势。
A four-point moving average for quarterly data is calculated by averaging each set of four consecutive values. For annual data, a three-point moving average is common. If the moving average increases steadily, this suggests a positive trend in regional integration.
对于季度数据，四点移动平均通过计算每四个连续值的平均值来获得。对于年度数据，三点移动平均较为常见。如果移动平均稳定上升，这表明区域一体化呈正趋势。
- Year 1: 45, Year 2: 50, Year 3: 56, Year 4: 60
- 第 1 年：45，第 2 年：50，第 3 年：56，第 4 年：60
- Three-point moving averages: (45+50+56)/3 = 50.3, then (50+56+60)/3 = 55.3
- 三点移动平均：(45+50+56)/3 = 50.3，然后是 (50+56+60)/3 = 55.3
10. Chi-Squared Tests for Regional Policy Preferences | 区域政策偏好的卡方检验
We can use a chi-squared test for independence to investigate whether support for deeper EU integration is associated with region type, such as urban, rural, or industrial. The data are arranged in a contingency table showing observed frequencies for each category. The expected frequencies are calculated under the assumption that the variables are independent.
我们可以使用卡方独立性检验来研究对深化欧盟一体化的支持是否与区域类型（如城市、农村或工业区）相关。数据排列在列联表中，显示每个类别的观测频数。期望频数是在变量独立的假设下计算的。
Expected frequency = (row total × column total) / grand total
χ² = Σ (O − E)² / E , degrees of freedom = (r − 1)(c − 1)
If the calculated chi-squared statistic exceeds the critical value at the chosen significance level, we reject the null hypothesis of independence and conclude that regional type and policy preference are associated.
如果计算出的卡方统计量超过所选显著性水平下的临界值，我们就拒绝独立性原假设，并得出结论：区域类型与政策偏好之间存在关联。
| Region type | Support integration | Oppose integration | Row total |
|---|---|---|---|
| Urban | 120 | 30 | 150 |
| Rural | 60 | 40 | 100 |
| Industrial | 90 | 60 | 150 |
| Column total | 270 | 130 | 400 |
. Repeating this for all cells allows us to compute the chi-squared statistic and compare it with the critical value.
对于城市/支持单元格，期望频数 = (150 × 270) / 400 = 101.25。对所有单元格重复此计算，我们就可以计算卡方统计量并将其与临界值进行比较。
11. Binomial and Poisson Models for EU Funding Allocation | 欧盟资金分配的二项与泊松模型
The number of regions receiving a specific type of EU funding in a given year can be modelled using a binomial distribution if we have a fixed number of regions and a constant probability of receiving funding. If the probability is small and the number of regions is large, the Poisson distribution can be used as an approximation.
如果区域数量固定且获得资金的概率恒定，则某一年获得特定类型欧盟资金的区域数量可以用二项分布建模。如果概率很小而区域数量很大，则可以用泊松分布作为近似。
X ~ B(n, p), P(X = r) = ⁿCᵣ pʳ (1 − p)ⁿ⁻ʳ
Poisson approximation: X ~ Po(np) when n is large and p is small
ty that exactly 10 regions receive funding is approximately P(X = 10) = (e⁻¹² × 12¹⁰) / 10! ≈ 0.104.
例如，如果 240 个区域每个获得特定创新拨款的概率为 0.05，那么受资助区域的期望数量为 240 × 0.05 = 12。恰好有 10 个区域获得资助的概率约为 P(X = 10) = (e⁻¹² × 12¹⁰) / 10! ≈ 0.104。
12. Limitations of Statistical Models in Regionalism | 区域主义统计模型的局限性
Statistical models simplify reality and rely on assumptions that may not hold in the complex political and economic context of the EU. For example, the assumption of independence in Poisson or binomial models may be violated if regions influence each other through spillover effects. Regression models may suffer from omitted variable bias if important factors like institutional quality or historical ties are not included.
统计模型简化了现实，并依赖于在欧盟复杂的政治经济背景下可能不成立的假设。例如，如果区域之间通过溢出效应相互影响，泊松或二项模型中的独立性假设可能被违反。如果未包括制度质量或历史联系等重要因素，回归模型可能存在遗漏变量偏差。
Furthermore, correlation is not causation, and any observed relationship between regionalism indicators and economic outcomes should be interpreted with caution. Good statistical practice requires checking residuals, identifying outliers, and using hypothesis tests to support conclusions rather than relying on a single measure.
此外，相关性不是因果关系，观察到的区域主义指标与经济结果之间的任何关系都应谨慎解释。良好的统计实践要求检查残差、识别异常值，并使用假设检验来支持结论，而不是依赖单一度量。
Published by TutorHao | Maths 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