📚 Israel and Italy: A Data-Driven Exploration for IB Math | 以色列与意大利：IB数学中的数据驱动探究
In the IB Mathematics curriculum, students are encouraged to apply statistical techniques and mathematical modelling to real-world contexts. A comparative investigation of two nations — Israel and Italy — provides a rich opportunity to explore descriptive statistics, hypothesis testing, correlation, and regression. This article walks you through a structured data analysis that mirrors the Internal Assessment (IA) process, using key indicators such as GDP per capita, life expectancy, and population growth.
在 IB 数学课程中，学生需要将统计技巧和数学模型应用于真实情境。对以色列和意大利两个国家进行比较探究，为探索描述性统计、假设检验、相关性与回归提供了丰富的素材。本文将以内部评估（IA）的流程为框架，利用人均 GDP、预期寿命和人口增长等关键指标，一步步展开数据分析。
1. Introduction: Why Compare Israel and Italy? | 引言：为何比较以色列与意大利？
Israel and Italy sit at opposite ends of the Mediterranean, yet they share similarities in population size (around 9 million for Israel, 59 million for Italy) and advanced economies. Comparing them allows IB students to investigate whether significant differences exist in living standards or demographic trends, and to practise statistical inference on real data. The guiding question could be: 'Is there a statistically significant difference between the mean life expectancy of Israel and Italy?'
以色列和意大利分属地中海两端，人口规模有差异（以色列约 900 万，意大利约 5900 万），但同属发达经济体。比较它们可以帮助 IB 学生探究生活水平或人口趋势是否存在显著差异，并在真实数据上练习统计推断。核心问题可以是：“以色列与意大利的平均预期寿命是否存在统计显著差异？”
2. Data Collection: Key Indicators | 数据收集：关键指标
For a robust investigation, we collect annual data for the past 30 years from sources like the World Bank. Indicators include GDP per capita (current US$), life expectancy at birth (years), population growth rate (%), and the Human Development Index (HDI). A sample of the dataset is shown below.
为保证探究的可靠性，我们从世界银行等来源收集了过去 30 年的年度数据，指标包括人均 GDP（现价美元）、出生时预期寿命（岁）、人口增长率（%）以及人类发展指数（HDI）。部分数据样本如下表所示。
| Year | Country | GDP per capita (US$) | Life expectancy (years) | Population growth (%) |
|---|---|---|---|---|
| 2020 | Israel | 44,169 | 82.8 | 1.7 |
| 2020 | Italy | 31,714 | 83.2 | -0.4 |
| 2015 | Israel | 36,051 | 82.0 | 1.9 |
| 2015 | Italy | 29,958 | 82.9 | -0.2 |
3. Descriptive Statistics: Mean, Median, Standard Deviation | 描述性统计：均值、中位数、标准差
Before performing any test, we summarise the life expectancy data for each country. Let x̄₁ and s₁ denote the sample mean and standard deviation for Israel, and x̄₂, s₂ for Italy. From the collected 30-year data, we obtain:
在进行任何检验之前，我们先对两国的预期寿命数据进行汇总。令 x̄₁ 和 s₁ 代表以色列的样本均值和标准差，x̄₂ 和 s₂ 代表意大利的。根据收集的 30 年数据，我们得到：
Israel: x̄₁ = 81.2 years, s₁ = 1.5 years, n₁ = 30
Italy: x̄₂ = 82.5 years, s₂ = 1.2 years, n₂ = 30
The median life expectancy for Israel is 81.0 years, slightly lower than the mean, while Italy's median is 82.7 years. The standard deviation indicates that life expectancy in Israel varies a bit more over the period studied compared to Italy.
以色列预期寿命中位数为 81.0 岁，略低于均值；意大利中位数为 82.7 岁。标准差显示，在研究时段内以色列的预期寿命波动略大于意大利。
4. Visualising Data: Box Plots and Histograms | 数据可视化：箱线图与直方图
A side-by-side box plot reveals that Italy's life expectancy distribution is shifted slightly higher, with less spread. Both distributions appear roughly symmetric, though Israel shows one mild low outlier in the early 1990s. Histograms confirm a unimodal shape, supporting the assumption of approximate normality for each group.
并排箱线图显示，意大利预期寿命的整体分布稍高且更集中。两国的分布大致对称，不过以色列在 20 世纪 90 年代初有一个温和的低异常值。直方图证实了单峰形态，支持每组数据近似正态的假设。
5. Comparing Means: Two-Sample t-Test | 均值比较：双样本 t 检验
answer our guiding question, we perform a two-tailed, independent-samples t-test at the 5% significance level. The null hypothesis is H₀: µ₁ = µ₂ (no difference in population means), and the alternative is H₁: µ₁ ≠ µ₂.
为回答核心问题，我们在 5% 显著性水平下进行双尾独立样本 t 检验。原假设为 H₀: µ₁ = µ₂（总体均值无差异），备择假设为 H₁: µ₁ ≠ µ₂。
The test statistic is calculated as:
检验统计量计算如下：
t = (x̄₁ - x̄₂) / √s₁²/n₁ + s₂²/n₂ = (81.2 - 82.5) / √1.5²/30 + 1.2²/30 ≈ -3.71
n₂-1, i.e., 29), the critical value for α = 0.05 two-tailed is approximately ±2.045. Since |-3.71| cally significant evidence that the mean life expectancy differs between Israel and Italy.
采用保守自由度（取 n₁-1 与 n₂-1 中较小值，即 29），α = 0.05 双尾检验的临界值约为 ±2.045。由于 |-3.71| > 2.045，我们拒绝原假设。有统计显著证据表明，以色列和意大利的平均预期寿命存在差异。
6. Assumptions and Conditions | 假设与条件
9,29) at α = 0.05, so we do not reject equal variances. Thus the standard t-test is appropriate.
t 检验的有效性依赖于独立性、近似正态性以及方差相等或可比。我们的数据点是独立的年度观测值，直方图未出现严重偏斜。方差齐性的 F 检验给出 F = s₁²/s₂² ≈ 1.56，低于 α = 0.05 下 F(29,29) 的临界值，所以不拒绝方差相等的假设。因此，标准 t 检验是合适的。
7. Correlation Analysis: GDP per Capita vs Life Expectancy | 相关性分析：人均 GDP 与预期寿命
We now examine the relationship between GDP per capita and life expectancy across both countries combined. Plotting the data reveals a positive association. The Pearson correlation coefficient r is calculated as 0.68, indicating a moderately strong linear relationship.
现在我们来考察两国人均 GDP 与预期寿命之间的关系。绘制散点图可看到正相关。计算得到皮尔逊相关系数 r = 0.68，表明存在中等强度的线性关系。
r Israel suggests that economic growth has been more tightly coupled with health improvements in recent decades.
仅以色列的数据 r = 0.74；意大利为 r = 0.59。以色列更强的相关性表明，近几十年来其经济增长与健康改善的关联更为紧密。
8. Regression Model and Prediction | 回归模型与预测
Using least-squares regression on the pooled data (n = 60), we obtain the linear model:
利用合并数据（n = 60）进行最小二乘回归，得到线性模型：
Life Expectancy = 70.2 + 0.00028 × GDP per capita
The slope 0.00028 means that an increase of US$ 28 years in life expectancy, on average. The coefficient of determination R² = 0.46, so 46% of the variation in life expectancy is explained by GDP per capita.
斜率 0.00028 表示人均 GDP 每增加 1,000 美元，预期寿命平均增加 0.28 岁。决定系数 R² = 0.46，即预期寿命 46% 的变异可由人均 GDP 解释。
Using this model, if Israel's GDP per capita reaches US$ Such extrapolation should be treated with caution, as the relationship may not remain linear beyond the observed range.
根据该模型，若以色列人均 GDP 达到 50,000 美元，预期寿命预测值为 70.2 + 0.00028×50000 = 84.2 岁。这种外推需谨慎对待，因为在观测范围之外关系可能不再保持线性。
9. Limitations and Validity | 局限性与有效性
The analysis is limited by the use of aggregated national data, which may mask within-country inequalities. The 30-year time span could introduce serial correlation, though annual observations reduce this risk. Additionally, confounding variables such as healthcare expenditure, diet, and education levels are not controlled for, so causality cannot be inferred from correlation alone.
本分析受限于使用全国汇总数据，这可能掩盖国内的不平等。30 年的时间跨度可能引入序列相关，但采用年度观测值降低了这一风险。此外，医疗支出、饮食、教育水平等混杂变量未加以控制，因此不能仅凭相关性推断因果关系。
10. Conclusion and IA Tips | 结论与内部评估建议
This investigation demonstrates a full IB-style data analysis cycle: formulating a question, collecting data, performing descriptive and inferential statistics, and discussing limitations. Students can extend this by incorporating more countries, using non-parametric tests, or exploring multiple regression. Remember to cite all data sources, label graphs clearly, and reflect on methodological choices to score highly on Criterion E (Reflection).
本探究展示了完整的 IB 风格数据分析流程：提出问题、收集数据、进行描述性和推断性统计，并讨论局限性。学生可以进一步拓展，加入更多国家、使用非参数检验或探索多元回归。请记住引用所有数据来源、清晰标注图表，并对方法选择进行反思，以便在标准 E（反思）中获得高分。
Published by TutorHao | Mathematics 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