📚 Year 12 SQA Statistics: In-Depth Analysis of Past Papers | SQA 统计历年真题深度解析
The SQA Higher Statistics course challenges students to apply rigorous statistical reasoning to real-world data. A careful study of past papers reveals recurring question types, common pitfalls, and the depth of understanding required to achieve top marks. This in-depth revision guide dissects key topics using authentic exam-style examples, providing you with bilingual explanations and proven strategies to master the assessment.
SQA 高等统计课程要求学生将严谨的统计推理应用于现实数据。仔细研究历年真题能够揭示反复出现的题型、常见易错点以及取得高分所需的理解深度。这份深度复习指南利用真实的考试风格例题剖析核心主题，为你提供双语解释和经过验证的策略，助力你攻克评估。
1. Course Structure and Mark Allocation | 课程结构与分值分配
The SQA Higher Statistics exam typically consists of two papers. Paper 1, non-calculator, tests foundational skills and theoretical understanding, while Paper 2 allows the use of a calculator and focuses on applied problem-solving. Marks are distributed across descriptive statistics, probability, distributions, inference, and regression.
SQA 高等统计考试通常由两份试卷组成。试卷一不允许使用计算器，测试基本技能和理论理解；试卷二允许使用计算器，侧重应用问题解决。分值分布在描述性统计、概率、分布、推断和回归等模块。
Past papers show that approximately 40% of marks relate to probability and distributions, 30% to statistical inference, 20% to data analysis, and 10% to evaluation and interpretation. Command words such as ‘calculate’, ‘interpret’, ‘comment on’, and ‘justify’ appear frequently.
历年真题显示，约40% 的分值与概率和分布相关，30%与统计推断相关，20%与数据分析相关，10%与评价和解释相关。高频出现的指令词包括“计算”、“解释”、“评论”和“论证”。
2. Descriptive Statistics and Data Presentation | 描述性统计与数据呈现
Many Paper 1 questions require calculation of the mean, median, interquartile range and standard deviation from a small data set. Understanding the effect of adding or multiplying constants is also examined.
试卷一的许多题目要求从小数据集中计算平均数、中位数、四分位距和标准差。同时也会考查增加或乘以常数的影响。
Worked example: The masses of six apples (g) are 142, 151, 148, 139, 155, 147. Find the mean and standard deviation. 解题示例：六个苹果的质量（克）为142, 151, 148, 139, 155, 147。求平均数和标准差。
To compute the mean, sum all values: 142+151+148+139+155+147 = 882. Divide by n=6 to obtain x̄ = 147 g.
计算平均数时，首先求和：142+151+148+139+155+147 = 882。除以 n=6 得到 x̄ = 147 g。
The sample standard deviation is found using the formula s = √[ Σ(xᵢ – x̄)² / (n-1) ]. Deviations are -5, 4, 1, -8, 8, 0. Squared deviations: 25, 16, 1, 64, 64, 0. Sum = 170. Divide by 5 gives 34. Hence s = √34 ≈ 5.83 g.
样本标准差使用公式 s = √[ Σ(xᵢ – x̄)² / (n-1) ] 计算。离差分别为 -5, 4, 1, -8, 8, 0。平方后为25, 16, 1, 64, 64, 0。总和为170。除以5得34。因此 s = √34 ≈ 5.83 g。
A common exam question asks you to draw a boxplot for given data or compare two boxplots, commenting on centre, spread and skewness. Always label axes clearly and provide a comparative statement supported by statistics.
常见的考题要求为给定数据绘制箱线图，或比较两个箱线图，评论其中心、分散程度和偏度。务必清晰标注坐标轴，并用统计量支撑比较性陈述。
3. Probability, Venn Diagrams and Tree Diagrams | 概率、维恩图和树状图
Probability questions often involve conditional probabilities and the use of Venn diagrams to represent events. Past papers test the ability to translate a word problem into set notation and calculate P(A|B) = P(A ∩ B) / P(B).
概率题常涉及条件概率以及使用维恩图表示事件。历年真题考查将文字问题转化为集合符号并计算 P(A|B) = P(A ∩ B) / P(B) 的能力。
Consider a past paper context: 60% of students study Mathematics (M), 45% study Statistics (S), and 25% study both. Complete a Venn diagram and find the probability that a randomly chosen student studies Mathematics given that they study Statistics.
考虑真题情境：60% 的学生学习数学(M)，45% 学习统计(S)，25% 两者都学。补全维恩图，并求随机选出一名学生在学习统计的条件下学习数学的概率。
The intersection is 0.25. The region for M only is 0.60 – 0.25 = 0.35; S only is 0.45 – 0.25 = 0.20. Outside both is 1 – (0.35+0.25+0.20) = 0.20. Then P(M|S) = 0.25 / 0.45 = 5/9.
交集部分为0.25。仅 M 的区域为0.60 – 0.25 = 0.35；仅 S 的区域为0.45 – 0.25 = 0.20。两者都不学的概率为1 – (0.35+0.25+0.20) = 0.20。因此 P(M|S) = 0.25 / 0.45 = 5/9。
Tree diagrams are frequently used for successive events. Be precise with branching probabilities and highlight paths for compound events. Always check if events are independent by verifying P(A) × P(B) = P(A ∩ B).
树状图常用于连续事件。对于分支概率要精确，并突出复合事件的路径。始终通过验证 P(A) × P(B) = P(A ∩ B) 来检查事件是否独立。
4. Discrete Random Variables and Binomial Distribution | 离散随机变量与二项分布
The binomial distribution X ~ B(n, p) is a cornerstone of the Higher Statistics course. You must be able to calculate probabilities using the formula P(X = k) = ⁿCₖ pᵏ (1-p)ⁿ⁻ᵏ, find expected value E(X) = np and variance Var(X) = np(1-p).
二项分布 X ~ B(n, p) 是高等统计课程的基石。你必须能够使用公式 P(X = k) = ⁿCₖ pᵏ (1-p)ⁿ⁻ᵏ 计算概率，会求期望 E(X) = np 和方差 Var(X) = np(1-p)。
A typical exam question: A biased coin lands heads with probability 0.4. It is tossed 10 times. Find the probability of exactly 4 heads and the probability of at least 8 heads.
典型考题：一枚不均匀硬币出现正面的概率为0.4。抛掷10次。求恰好出现4次正面的概率，以及至少出现8次正面的概率。
Using the formula: P(X = 4) = C(10,4) × (0.4)⁴ × (0.6)⁶. C(10,4) = 210, (0.4)⁴ = 0.0256, (0.6)⁶ ≈ 0.04666. So P ≈ 210 × 0.0256 × 0.04666 ≈ 0.2508.
代入公式：P(X = 4) = C(10,4) × (0.4)⁴ × (0.6)⁶。C(10,4) = 210，(0.4)⁴ = 0.0256，(0.6)⁶ ≈ 0.04666。因此概率约等于 210 × 0.0256 × 0.04666 ≈ 0.2508。
For at least 8 heads, sum P(8) + P(9) + P(10). Compute each: P(8) ≈ 0.0106, P(9) ≈ 0.0016, P(10) ≈ 0.0001, giving a total of about 0.0123. This shows the distribution is highly skewed when p ≠ 0.5.
对于至少8次正面，累加 P(8) + P(9) + P(10)。分别计算：P(8) ≈ 0.0106，P(9) ≈ 0.0016，P(10) ≈ 0.0001，总和约为0.0123。这表明当 p ≠ 0.5 时分布高度偏斜。
Be ready to use binomial tables or calculator functions efficiently in Paper 2. In Paper 1, manual calculation of small n and k is common.
在试卷二中，熟练使用二项分布表或计算器功能。试卷一则常见对较小 n 和 k 的手动计算。
5. Normal Distribution and Standardisation | 正态分布与标准化
The normal distribution N(μ, σ²) is used extensively. Questions ask you to find probabilities, critical values, or unknown μ and σ. Standardisation using Z = (X – μ)/σ transforms any normal variable to N(0,1).
正态分布 N(μ, σ²) 被广泛使用。问题要求你计算概率、临界值，或未知的 μ 和 σ。通过标准化 Z = (X – μ)/σ 可将任意正态变量转化为 N(0,1)。
Example from a recent paper: The weights of a breed of dogs are normally distributed with mean 28 kg and standard deviation 4 kg. Find the proportion of dogs weighing between 25 kg and 32 kg.
近期真题示例：某犬种的体重服从正态分布，均值为28 kg，标准差为4 kg。求体重介于25 kg与32 kg之间的比例。
Standardise the boundaries: Z₁ = (25 – 28)/4 = -0.75, Z₂ = (32 – 28)/4 = 1.00. Use the standard normal table: Φ(1.00) = 0.8413, Φ(-0.75) = 0.2266. The area between is 0.8413 – 0.2266 = 0.6147. About 61.5% of dogs fall in this range.
对边界标准化：Z₁ = (25 – 28)/4 = -0.75，Z₂ = (32 – 28)/4 = 1.00。查标准正态表：Φ(1.00) = 0.8413，Φ(-0.75) = 0.2266。中间面积为0.8413 – 0.2266 = 0.6147。约有61.5% 的犬只属于这个范围。
Inverse normal problems require finding X given a probability. Sketch a curve, shade the region, and use the percentile Z-value to solve X = μ + Zσ. Past papers often embed this within hypothesis testing or confidence intervals.
反向正态问题需要根据概率求 X。绘制曲线，标示阴影区域，并使用百分位数 Z 值代入 X = μ + Zσ 求解。真题常将此嵌入在假设检验或置信区间之中。
6. Sampling and the Central Limit Theorem | 抽样与中心极限定理
The Central Limit Theorem (CLT) states that for a sufficiently large sample size (n ≥ 30), the sampling distribution of the mean is approximately normal with mean μ and standard error σ/√n, regardless of the population’s shape.
中心极限定理指出，当样本量足够大（n ≥ 30）时，样本均值的抽样分布近似服从均值为 μ、标准误为 σ/√n 的正态分布，无论总体形状如何。
Exam questions frequently provide a population mean and standard deviation, then ask for the probability that a sample mean exceeds a certain value. For instance, the time spent on a revision website is skewed with μ = 45 min, σ = 20 min. Find P(x̄ > 50) for a sample of 36 students.
考题常给出总体均值和标准差，然后要求计算样本均值超过某个值的概率。例如，学生在某复习网站的停留时间呈偏态分布，μ = 45分钟，σ = 20分钟。求样本量为36的学生平均停留时间大于50分钟的概率。
By CLT, x̄ ~ N(45, 20²/36) i.e. N(45, (3.33..)²). Then Z = (50 – 45) / (20/√36) = 5 / 3.333.. = 1.5. P(Z > 1.5) = 1 – 0.9332 = 0.0668.
根据中心极限定理，x̄ 服从 N(45, 20²/36)，即 N(45, (3.33..)²)。Z = (50 – 45) / (20/√36) = 5 / 3.333.. = 1.5。P(Z > 1.5) = 1 – 0.9332 = 0.0668。
Remember to state the CLT justification explicitly in your solution. Examiners award marks for recognising that the distribution of sample means tends to normality.
务必在解答中明确陈述中心极限定理的依据。阅卷人会因为你识别出样本均值分布趋于正态而给分。
7. Confidence Intervals for the Mean | 均值的置信区间
A confidence interval provides a range of plausible values for a population parameter. The 95% symmetric confidence interval for μ when σ is known is x̄ ± z* × σ/√n, where z* = 1.96.
置信区间为总体参数提供了一个合理的取值范围。当 σ 已知时，μ 的95%对称置信区间为 x̄ ± z* × σ/√n，其中 z* = 1.96。
A past paper question gave sample mean mass 152 g, population standard deviation 8 g, and sample size 25. Construct a 95% confidence interval and interpret it.
一道真题给出样本均值152 g，总体标准差8 g，样本量25。构建95%置信区间并加以解释。
Standard error = 8/√25 = 1.6. Margin of error = 1.96 × 1.6 = 3.136. So the interval is (152 – 3.136, 152 + 3.136) = (148.86, 155.14). We are 95% confident that the true population mean mass lies between 148.86 g and 155.14 g.
标准误 = 8/√25 = 1.6。误差范围 = 1.96 × 1.6 = 3.136。因此区间为 (152 – 3.136, 152 + 3.136) = (148.86, 155.14)。我们有95%的把握认为真实的总体平均质量介于148.86克到155.14克之间。
When σ is unknown, the t-distribution is used. Past papers test the ability to find t* from tables using n-1 degrees of freedom, especially for small samples. The interpretation must include the confidence level and the parameter in context.
当 σ 未知时，使用 t 分布。真题测试根据自由度 n-1 查表找到 t* 的能力，尤其针对小样本。解释时必须结合背景注明置信水平和参数。
8. Hypothesis Testing: One-sample z-test | 假设检验：单样本 z 检验
Hypothesis testing is a key inferential procedure. Questions require stating null (H₀: μ = μ₀) and alternative (H₁: μ ≠ μ₀ or one-sided) hypotheses, calculating the test statistic, comparing with critical value, and writing a conclusion in context.
假设检验是一项关键的推断程序。题目要求陈述原假设（H₀: μ = μ₀）和备择假设（H₁: μ ≠ μ₀ 或单侧），计算检验统计量，与临界值比较，并结合背景撰写结论。
Example: A manufacturer claims bolts have mean length 50 mm. A sample of 40 bolts gives x̄ = 50.7 mm, with known σ = 2 mm. Test at the 5% significance level whether the mean has changed.
示例：某制造商声称螺栓平均长度为50毫米。随机抽取40个螺栓，得到 x̄ = 50.7 mm，已知 σ = 2 mm。在5%显著性水平下检验均值是否发生变化。
Hypotheses: H₀: μ = 50, H₁: μ ≠ 50. Test statistic: Z = (50.7 – 50) / (2/√40) = 0.7 / 0.3162 ≈ 2.21. Critical value for 5% two-tailed is ±1.96. Since 2.21 > 1.96, reject H₀. There is sufficient evidence at the 5% level to suggest the mean length is not 50 mm.
假设：H₀: μ = 50，H₁: μ ≠ 50。检验统计量：Z = (50.7 – 50) / (2/√40) = 0.7 / 0.3162 ≈ 2.21。5% 双尾临界值为 ±1.96。由于 2.21 > 1.96，拒绝 H₀。在5%水平下有充分证据表明平均长度不是50毫米。
In Paper 1, p-values are sometimes required. Report the p-value and compare with significance level. If p-value < 0.05, reject H₀. Always end with a non-technical summary sentence.
在试卷一中，某些题目要求计算 p 值。报告 p 值并与显著性水平比较。若 p 值小于0.05，则拒绝 H₀。结尾总要用一句非技术性的总结陈述。
9. Chi-squared Test for Association | 卡方独立性检验
The chi-squared test assesses whether two categorical variables are associated. Past papers provide contingency tables and ask you to calculate expected frequencies, the test statistic, degrees of freedom, and state a conclusion.
卡方检验用于评估两个分类变量是否相关。真题提供列联表，要求计算期望频数、检验统计量、自由度并给出结论。
Expected frequencies are computed as (row total × column total) / grand total. The χ² statistic is Σ (O – E)² / E. Degrees of freedom = (rows – 1)×(columns – 1).
期望频数计算公式为（行合计 × 列合计）/ 总计。χ² 统计量为 Σ (O – E)² / E。自由度 = (行数 – 1) × (列数 – 1)。
A typical question might involve a survey of favourite sport by gender. Suppose χ² calculated is 6.45 and critical value at 5% with df=2 is 5.991. Since 6.45 > 5.991, reject H₀ that sport preference and gender are independent. There is evidence of an association.
典型题目可能涉及按性别划分的最喜爱运动调查。假设计算得 χ² = 6.45，5% 水平下自由度为2的临界值为5.991。由于6.45 > 5.991，拒绝运动偏好与性别独立的原假设。有证据表明二者存在关联。
Be careful to combine categories if expected frequencies fall below 5. Marks are also allocated for stating the conclusions in the context of the problem.
若期望频数低于5，需注意合并类别。结合问题背景陈述结论也可获得分数。
10. Correlation and Linear Regression | 相关与线性回归
Product-moment correlation coefficient r measures the strength and direction of a linear relationship. Its value lies between -1 and 1. Past exams often give summations Σx, Σy, Σx², Σy², Σxy and ask you to compute r.
积矩相关系数 r 衡量线性关系的强度和方向。其取值范围在 -1 到 1 之间。历年考试常给出 Σx, Σy, Σx², Σy², Σxy 的求和，要求计算 r。
Formula for r: r = Sxy / √(Sxx × Syy) where Sxx = Σx² – (Σx)²/n, Syy = Σy² – (Σy)²/n, Sxy = Σxy – (Σx Σy)/n. Interpret the value, e.g. r = 0.82 indicates a strong positive linear association.
r 的公式：r = Sxy / √(Sxx × Syy)，其中 Sxx = Σx² – (Σx)²/n，Syy = Σy² – (Σy)²/n，Sxy = Σxy – (Σx Σy)/n。对数值进行解释，例如 r = 0.82 表示强正线性相关。
Linear regression equation takes the form y = a + bx, where b = Sxy / Sxx and a = ȳ – b x̄. Use this for prediction within the range of data. Extrapolation is unreliable and should be noted.
线性回归方程形如 y = a + bx，其中 b = Sxy / Sxx，a = ȳ – b x̄。可用其对数据范围内的值进行预测。外推法不可靠，应予以说明。
In past papers, a follow-up question asks: ‘Comment on the reliability of using the regression line to predict for an x-value far outside the data set.’ The correct response discusses the dangers of extrapolation.
在真题中，后续问题会问：“对使用回归线预测远超数据集范围的 x 值，评论其可靠性。”正确的回答应讨论外推法的危险性。
11. Exam Technique: Command Words and Marking Schemes | 考试技巧：指令词与评分标准
Understanding command words is crucial. ‘Calculate’ requires a numerical answer with supporting working. ‘Interpret’ demands a contextual sentence linking the statistics to the scenario. ‘Justify’ or ‘Comment on’ expects a reasoned argument, often referring to a comparison of values or checking assumptions.
理解指令词至关重要。“计算”要求给出数值答案并附有计算过程。“解释”需要将统计量与情境关联起来写一句上下文相关的句子。“论证”或“评论”则期望一个有理有据的分析，常涉及对比数值或检查假设。
Examiners’ reports highlight that students often lose marks by omitting units, not showing substitution steps, or failing to state a conclusion in hypothesis testing. Always write a final sentence like ‘There is insufficient evidence to reject H₀’ rather than just ‘accept H₀’.
考官报告指出，学生常因遗漏单位、未展示代入步骤或在假设检验中未陈述结论而失分。务必写出如“没有充分证据拒绝 H₀”这样的结束句，而不是简单写“接受 H₀”。
When comparing boxplots, mention median, IQR, range, and skewness explicitly. For probability, use proper notation and always check that probabilities sum to 1. A well-structured answer with clear labels and logical flow will maximise marks.
比较箱线图时，要明确提及中位数、四分位距、全距和偏度。对于概率，使用正确的符号并始终检查概率之和是否为1。结构清晰、标签明确且逻辑流畅的答案将最大化得分。
Published by TutorHao | Statistics Revision Series | aleveler.com
更多咨询请联系16621398022（同微信）
屏轩国际教育cambridge primary/secondary checkpoint, cat4, ukiset,ukcat,igcse,alevel,PAT,STEP,MAT, ibdp,ap,ssat,sat,sat2课程辅导，国外大学本科硕士研究生博士课程论文辅导