📚 Hypothesis Testing for Zero Correlation | 零相关假设检验
In the Edexcel A-Level Mathematics specification, hypothesis testing for zero correlation is used to determine whether a calculated product moment correlation coefficient from a sample provides enough evidence to conclude that two variables are linearly related in the population.
在 Edexcel A-Level 数学考试中，零相关假设检验用于判断样本计算出的积矩相关系数是否提供足够证据，从而推断两个变量在总体中具有线性关系。
1. The Product Moment Correlation Coefficient (PMCC) | 积矩相关系数 (PMCC)
The product moment correlation coefficient, usually denoted by r, measures the strength and direction of the linear relationship between two variables. Its value always lies between -1 and 1.
积矩相关系数，通常记作 r，用于衡量两个变量之间线性关系的强度和方向。它的取值范围始终在 -1 到 1 之间。
while r = 1 or r = -1 indicates a perfect positive or perfect negative linear correlation.
r = 0 表示没有线性相关，而 r = 1 或 r = -1 表示完全正相关或完全负相关。
0 in the population, a sample may still produce a non-zero r because of random variation.
在假设检验中，r 被视为样本统计量，用于估计总体相关系数 ρ（希腊字母 rho）。即使总体中 ρ = 0，样本也可能因为随机波动而产生非零的 r。
The formula for the PMCC from n pairs of data is:
由 n 对数据计算 PMCC 的公式为：
r = Σ(xᵢ - x̄)(yᵢ - ȳ) / √(Σ(xᵢ - x̄)² × Σ(yᵢ - ȳ)²)
Here x̄ and ȳ are the sample means of the x and y values. The numerator measures how the two variables vary together, while the denominator standardises this by the individual spreads.
其中 x̄ 和 ȳ 分别是 x 和 y 的样本均值。分子衡量两个变量共同变化的程度，分母则用各自的离散程度进行标准化。
2. Why Test for Zero Correlation? | 为什么要检验零相关？
The phrase "zero correlation" de whether an observed sample correlation is large enough to be considered statistically significant, rather than simply due to sampling error.
"零相关" 具体指原假设 H₀ : ρ = 0。该检验帮助我们判断观察到的样本相关系数是否足够大，从而具有统计显著性，而不是仅仅由抽样误差引起。
a moderate positive relationship, but with only 12 points it may not be strong enough to conclude that the population correlation is non-zero.
例如，一个包含 12 对数据的样本可能得到 r = 0.43。这看起来是中等正相关，但在只有 12 个数据点的情况下，它可能不足以证明总体相关不为零。
Sample size matters greatly: small samples can produce surprisingly large r values by chance, so a formal significance test protects against over-interpreting weak evidence.
样本量非常重要：小样本可能偶然产生相当大的 r 值，因此正式的显著性检验可以防止过度解读微弱证据。
The test is widely used in contexts such as scientific experiments, business analytics, psychology and any situation where two measured variables may be linearly associated.
该检验广泛用于科学实验、商业分析、心理学以及任何两个测量变量可能存在线性关联的情境。
3. Stating Hypotheses for Zero Correlation | 零相关假设的陈述
The null hypothesis is always that there is no linear correlation in the population:
原假设始终为总体中不存在线性相关：
H₀ : ρ = 0
The alternative hypothesis can be two-tailed or one-tailed, depending on the wording of the question.
备择假设可以是双侧或单侧，具体取决于题目表述。
Two-tailed test for any linear correlation: H₁ : ρ ≠ 0
双侧检验，检验任何线性相关：H₁ : ρ ≠ 0
One-tailed test for positive correlation: H₁ : ρ > 0
单侧检验，检验正相关：H₁ : ρ > 0
One-tailed test for negative correlation: H₁ : ρ < 0
单侧检验，检验负相关：H₁ : ρ < 0
Words such as "there is a correlation", "there
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