📚 Measuring Secondary School Progression Outcomes: Key Indicators Explained | 中学阶段学生升学去向衡量指标说明
Every secondary school publishes something about where its leavers go, but the numbers behind those claims are rarely comparable. A “98% university progression rate” can mean anything from “98% of a hand-picked cohort of 40 applied and were accepted somewhere” to “98% of a full year group, verified by enrolment records, entered a degree programme within 18 months.” Understanding the underlying indicators – what they measure, what they exclude, and how they are calculated – is essential for students, parents, teachers and school leaders who want to make sound judgements rather than respond to marketing.
几乎每所中学都会公布毕业生的升学去向，但这些数字背后的口径往往并不可比。一个 “98% 升学率” 可能意味着 “40 人精选班级中有 98% 申请并被某处录取”，也可能意味着 “全年级 98% 的学生在 18 个月内凭注册记录被证实进入学位课程”。理解这些指标本身——它们衡量什么、排除什么、如何计算——对学生、家长、教师和学校管理者都至关重要，因为只有这样才能做出理性判断，而不是被宣传牵着走。
1. Why Progression Outcomes Matter | 为什么升学去向指标重要
Progression data serve at least four distinct purposes: accountability (did the school deliver on its educational promise?), school choice (which institution fits my goals?), institutional improvement (where are our counselling and academic gaps?), and equity monitoring (are outcomes distributed fairly across student groups?). Each purpose demands a slightly different indicator design, which is why a single headline percentage can never satisfy all four at once.
升学数据至少服务于四个不同目的：问责（学校是否兑现了教育承诺？）、择校（哪所学校更契合我的目标？）、学校改进（升学指导与教学的短板在哪里？）以及公平监测（不同学生群体的结果分布是否公正？）。每个目的都要求略有不同的指标设计，因此单一的核心百分比永远无法同时满足这四类需求。
A useful rule is that any indicator must answer three questions before it is quoted: Who is counted? What counts as a positive outcome? Over what time window? If a publication cannot answer all three, the figure should be treated as indicative rather than evidential.
一条实用原则是：任何指标在被引用前都必须回答三个问题——谁被计入？什么算作正面结果？在多长的时间窗口内？如果一份材料无法同时回答这三点，该数字只能当作参考性信息，而非证据。
2. Defining the Core Construct: Destination, Outcome and Progression | 核心概念界定：去向、结果与升学
These three terms are often used interchangeably, but they are analytically distinct. “Destination” is a descriptive statement of where a leaver is at a fixed point in time. “Outcome” is an evaluative judgement about the quality or durability of that destination. “Progression” describes movement into a next stage of education or training, usually with an implied standard of level or selectivity.
这三个词常被混用，但在分析上彼此有别。”去向” 是对毕业生在某一固定时点上所处位置的描述；”结果” 是对该去向质量或持续性的评价性判断；”升学” 则描述进入下一教育或培训阶段的流动，通常隐含对层次或选拔性的要求。
The time window is decisive. A snapshot taken three months after results day captures applications and conditional offers, but not enrolments, deferrals or clearing movements. An 18-month window captures gap years, re-sits and second-round applications, but introduces attrition from the sample. Most credible systems therefore report at two points: a short-term snapshot (6 months) and a medium-term verified figure (15 to 18 months).
时间窗口具有决定性意义。放榜后三个月的快照捕捉的是申请与有条件录取，而非实际注册、延期入学或补录流动；18 个月的窗口能覆盖间隔年、重考和第二轮申请，但会带来样本流失。因此，多数可靠体系会报告两个时点：短期快照（6 个月）与中期核实数据（15 至 18 个月）。
Progression Rate = (Number of leavers who progressed ÷ Total cohort size) × 100%
The denominator is where most disputes begin. Denominators in common use include the full year group, the subset who applied, those who applied by the main deadline, those who responded to the survey, or those with complete records. Only the full year group denominator supports genuine institutional comparison.
分歧几乎都从分母开始。常见的分母包括：全年级人数、提交申请者、在主轮截止前申请者、问卷作答者，或记录完整者。只有以全年级人数为分母，才支持真正意义上的学校间比较。
3. The Headline Indicator: University Progression Rate | 核心指标：大学升学率
The university progression rate is the most quoted and least standardised indicator in secondary education. Its meaning shifts entirely with three definitional choices: which institutions are counted, which qualification levels are counted, and whether the destination is verified.
大学升学率是中学阶段被引用最多、标准化程度最低的指标。它的含义完全取决于三个定义选择：计入哪些院校、计入哪些学历层次，以及去向是否经过核实。
| Variant | 变体 | Definition | 定义 | Typical effect | 典型影响 |
| Any HE | 任意高等教育 | Any post-secondary course of any level | Highest figure; weakest signal |
| Degree-level only | 仅学位层次 | Bachelor degree or equivalent only | Moderate; excludes foundation and diploma routes |
| Overseas only | 仅海外 | Institutions outside the home country | Inflated in systems with domestic capacity limits |
| Verified enrolment | 核实注册 | Confirmed by registry or enrolment evidence | Lowest but most defensible |
When comparing two schools, always request the variant. A school reporting “any HE, unverified, applicant denominator” is not producing a comparable statistic with one reporting “degree-level, verified enrolment, full cohort denominator”, even if both say 95%.
比较两所学校时，务必先索取口径说明。一所学校报告 “任意高教、未经核实、以申请者为分母”，与另一所报告 “学位层次、核实注册、全年级为分母”，即便都写作 95%，也不是可比统计量。
4. Destination Classification Systems | 去向分类体系
A classification system must be mutually exclusive and collectively exhaustive. Every leaver should fall into exactly one category, and the categories together must cover the whole cohort. The most robust systems include an explicit “unknown” or “not traced” category rather than absorbing non-respondents into a positive category.
分类体系必须做到互斥且穷尽：每位毕业生都应恰好落入一个类别，各类别合起来覆盖全体。最稳健的体系会明确设置 “未知” 或 “未追踪” 类别，而不是把未应答者并入正面类别。
- Higher education, degree level | 学位层次高等教育 – domestic or overseas, verified by enrolment.
- Higher education, sub-degree | 副学位层次高等教育 – foundation, diploma, associate routes.
- Apprenticeship or vocational training | 学徒制或职业培训 – often high-value but rarely counted in headline rates.
- Employment | 就业 – with a duration threshold, e.g. contracts of 6 months or more.
- Gap year or planned deferral | 间隔年或计划延期 – declared intent with evidence.
- Re-sit or repeat | 重考或复读 – full-time study aimed at reapplication.
- Not in education, employment or training | 未升学、未就业、未培训 – a legitimate and important category.
- Unknown or not traced | 未知或未追踪 – must be reported separately and never hidden.
A common malpractice is to report “university progression” as a percentage of those who applied, then publish it beside the full-cohort figure of a competitor. Another is to omit the unknown category entirely, which silently inflates every remaining percentage.
常见的失当做法是：以申请者为分母计算 “升学率”，再与竞争对手的全年级口径并列发布；或者干脆省略未知类别，从而静默推高所有剩余百分比。
5. Quality-Weighted Indicators: Selectivity and Tiering | 质量加权指标：选拔性与层次
A bare progression rate treats a highly selective research university and an open-access institution as equivalent. Quality-weighted indicators attempt to correct this by attaching a weight to each destination. The weights may reflect entry tariffs, admission selectivity, ranking bands or employer-recognised prestige.
单纯的升学率会把高度选拔性的研究型大学与开放入学的院校等同看待。质量加权指标试图通过给每个去向附加权重来修正这一点，权重可反映入学分数线、录取选拔性、排名区间或雇主认可度。
Quality-Weighted Progression Index = Σ (wᵢ × nᵢ) ÷ N
Here wᵢ is the weight assigned to destination category i, nᵢ is the number of leavers in that category, and N is the total cohort. The index is only as trustworthy as the published weighting scheme.
其中 wᵢ 为第 i 类去向的权重，nᵢ 为该类别的毕业生人数，N 为全体人数。该指数的可信度，完全取决于权重方案是否公开。
Three cautions apply. First, rankings are volatile, so weighting by exact rank position produces meaningless year-on-year swings; bands of five or ten places are more stable. Second, weightings embed value judgements about what counts as a good destination, which should be stated explicitly rather than presented as objective fact. Third, any weighting scheme should be tested by sensitivity analysis – if the school’s ranking changes materially when weights shift by 10%, the index is fragile.
有三点需注意。第一，排名波动大，按精确名次赋权会产生无意义的年度剧烈变化，五名或十名一档更稳定。第二，权重本身嵌入了 “什么是好去向” 的价值判断，应明确说明，而非包装成客观事实。第三，任何赋权方案都应做敏感性分析——如果权重变动 10% 就导致学校位次显著改变，该指数就是不稳健的。
6. Subject Fit and Course Alignment | 专业匹配度与课程契合度
Where students go matters, but so does what they study. Course alignment measures the proportion of leavers entering programmes that connect meaningfully to their prior specialisation, stated intentions or career trajectory. High alignment suggests effective guidance; low alignment may indicate indecision, mismatch or a weak counselling function.
学生去哪里重要，学什么同样重要。专业契合度衡量的是：进入与自身先前专长、既定意向或职业路径有意义关联的课程的学生比例。契合度高说明指导有效；契合度低可能反映方向不明、匹配失当或升学指导功能薄弱。
Alignment Rate = (Leavers entering a linked programme ÷ Total leavers entering HE) × 100%
Related indicators worth tracking include STEM progression share, progression into shortage or priority sectors, and the rate at which offers are converted into enrolments in the student’s first-choice subject area. In A-Level and IB contexts, alignment is typically assessed against the three or four subjects studied in the final two years.
值得一并追踪的相关指标包括：STEM 方向升学占比、进入紧缺或重点领域的比例，以及在首选专业方向上 “录取转为注册” 的转化率。在 A-Level 与 IB 语境中，契合度通常以最后两年所修的三到四门学科为参照进行评估。
7. Persistence, Completion and Withdrawal Rates | 持续就读、完成率与退学率
Entry is not success. A school may send many students to university and many may leave within a year. Persistence indicators measure whether the destination holds. They are usually reported by the receiving institution, which makes them less susceptible to school-level manipulation but harder to assemble for overseas cohorts.
入学不等于成功。一所学校可能送很多学生进大学，其中很多人却在一年内离开。持续性指标衡量的是去向是否稳固。这类数据通常由接收院校报告，因此较不易被中学层面操纵，但跨国统计的整合难度更高。
Retention Rate (Year 1 to Year 2) = (Still enrolled at start of Year 2 ÷ Enrolled at start of Year 1) × 100%
Useful companion measures are the first-year withdrawal rate, the course-transfer rate, the completion rate within the expected duration plus one year, and the proportion of students who change institution or course in the first 12 months. A school with a 95% progression rate and a 20% first-year withdrawal rate is not outperforming a school with 88% progression and 5% withdrawal; it is simply optimising for a different metric.
配套指标包括：一年级退学率、转专业率、在标准学制加一年内完成学位的比例，以及 12 个月内更换院校或课程的比例。一所升学率 95%、一年级退学率 20% 的学校，并不优于升学率 88%、退学率 5% 的学校，它只是针对不同指标做了优化。
8. Equity-Adjusted Progression: Gap Analysis | 公平调整后的升学指标：差距分析
Aggregate rates can conceal substantial internal inequality. Equity-adjusted indicators disaggregate outcomes by group – socioeconomic background, gender, ethnicity, special educational needs, first-generation status, or scholarship category – and report the gaps between them.
总体比率可能掩盖显著的内部不平等。公平调整后的指标会按群体拆分结果——社会经济背景、性别、族裔、特殊教育需求、家庭第一代大学生身份或奖学金类别——并报告群体之间的差距。
Equity Gap = P_disadvantaged group − P_reference group
Where P represents the progression rate for each group, expressed in percentage points. A negative gap indicates under-representation in progression; a gap near zero indicates parity. A single summary figure, the representation index, can be computed as the group’s share of high-selectivity destinations divided by its share of the cohort; a value of 1.0 indicates proportional representation.
其中 P 代表各群体的升学率，以百分点表示。负值表示升学上的代表性不足；接近零表示大致均衡。也可用单一汇总指标 “代表性指数”：某群体在高选拔性去向中的占比除以其在全年级中的占比；等于 1.0 表示比例相符。
Equity reporting should always state the size of each subgroup. Gaps computed on groups of fewer than 15 students are statistically unstable and should be suppressed or aggregated across years.
公平性报告必须同时说明各子群体的人数。少于 15 人的群体所计算出的差距在统计上不稳定，应当隐去或跨年度合并处理。
9. Value-Added and Contextual Value-Added Measures | 增值与情境化增值指标
Raw progression rates largely reflect intake. Selective schools with high prior attainment will always look strong. Value-added measures attempt to isolate the school’s contribution by comparing actual outcomes with those predicted from students’ prior attainment and background characteristics.
原始升学率在很大程度上反映的是生源。录取分数高的选拔性学校永远显得亮眼。增值指标试图通过将实际结果与依据学生先前学业水平和背景预测的结果相比较，来分离学校自身的贡献。
Value Added = Actual outcome − Predicted outcome
Contextual value-added (CVA) extends the model to include socioeconomic status, first language, mobility and special educational needs. A positive value indicates the school performed better than expected for its intake; a negative value indicates the reverse. Confidence intervals are essential: a value of +3 percentage points with an interval spanning −2 to +8 is statistically indistinguishable from zero.
情境化增值（CVA）进一步把社会经济地位、第一语言、学生流动性和特殊教育需求纳入模型。正值表示学校相对于其生源表现优于预期，负值则相反。置信区间不可或缺：+3 个百分点若区间跨 −2 至 +8，则与零在统计上无法区分。
Two design cautions: value-added models are sensitive to which prior-attainment measure is used, and small cohorts produce volatile values. A three-year rolling average is the standard remedy for both.
两点设计上的注意：增值模型对采用哪种先前学业指标十分敏感；小规模年级会产生剧烈波动。三年滚动平均是应对这两点的标准做法。
10. Data Sources, Coverage and Response Rates | 数据来源、覆盖度与回收率
Indicator quality is capped by data quality. The strongest systems triangulate multiple sources: admissions service records, receiving-institution enrolment confirmations, national student data returns, and school-level alumni tracking. Self-reported social media data is the weakest source, since those who post about university are systematically different from those who do not.
指标质量的上限由数据质量决定。最稳健的体系会交叉验证多个来源：招生系统记录、接收院校的注册确认、国家学生数据申报，以及学校层面的校友追踪。自我报告的社交媒体数据是最弱来源，因为在网上晒大学的人与不晒的人存在系统性差异。
Response Rate = (Valid responses ÷ Eligible cohort) × 100%
Response rate alone is insufficient; non-response bias matters more. If the 20% who do not respond are disproportionately the NEET and unknown groups, the reported progression rate is inflated regardless of how high the response rate is. Best practice is to report response rate, the composition of non-respondents where known, and the method used to verify claimed destinations.
仅有回收率并不够，无应答偏差更为关键。如果不作答的 20% 中 “未升学未就业” 和 “未知” 群体占比过高，那么无论回收率多高，报告出的升学率都会被高估。最佳实践是同时报告：回收率、已知无应答者的构成，以及核实所申报去向的方法。
11. Common Pitfalls and Misleading Comparisons | 常见误区与误导性比较
Most misleading statistics are not fabricated; they are selectively framed. Recognising the recurring traps is the fastest route to reading progression data competently.
大多数误导性统计并非捏造，而是选择性框定。识别反复出现的陷阱，是提升数据阅读能力最快的路径。
- Small denominators | 分母过小 – a cohort of 20 can swing 5 percentage points from a single student.
- Definitional drift | 定义漂移 – the same school using different definitions across years.
- Survivorship bias | 幸存者偏差 – publishing outcomes only for those who responded.
- Cherry-picked windows | 挑选时间窗口 – choosing the snapshot month that looks best.
- Ranking churn | 排名波动 – treating annual rank shifts as real quality changes.
- Gaming behaviour | 博弈行为 – discouraging weaker applicants from applying, to protect the rate.
- Category inflation | 类别注水 – counting sub-degree or short courses as degree progression.
- Unstated unknowns | 未披露的未知项 – the unknown category quietly disappearing from the table.
Cross-system comparison compounds every one of these problems. Comparing an A-Level cohort’s UCAS-based rates with an IB cohort’s global applications, or with a national system where university entry is centrally allocated by examination score, requires aligning not just the metric but the entire admissions mechanism.
跨体系比较会叠加上述所有问题。把 A-Level 群体基于 UCAS 的升学率，与 IB 群体的全球申请数据，或与按考试成绩统一分配学位的国家体系相比，需要对齐的不只是指标，而是整套招生机制。
12. A Balanced Scorecard and Practical Use | 均衡记分卡与实际应用
No single indicator deserves to stand alone. A defensible scorecard combines a volume measure, a quality measure, a durability measure and an equity measure, each with a declared denominator, time window and verification method.
没有任何单一指标应当独立支撑判断。一份站得住脚的记分卡应同时包含规模指标、质量指标、持续性指标和公平性指标，且每一指标都须声明分母、时间窗口与核实方式。
| Dimension | 维度 | Indicator | 指标 | Suggested weight | 建议权重 |
| Volume | 规模 | Degree-level progression rate, full cohort | 30% |
| Quality | 质量 | Quality-weighted index, banded weights | 25% |
| Durability | 持续性 | First-year retention rate | 20% |
| Equity | 公平 | Progression gap between subgroups | 15% |
| Data integrity | 数据完整性 | Verification rate, unknown category size | 10% |
When applying these indicators as a student or parent, ask four questions in this order: What is the denominator, and is it the full cohort? How were destinations verified, and what is the size of the unknown category? What time window was used, and is it consistent across years? How do outcomes differ across subgroups, and does the school publish them?
作为学生或家长使用这些指标时，请按以下顺序提出四个问题：分母是什么，是否为全年级？去向如何核实，未知类别规模多大？使用的时间窗口多长，各年度是否一致？不同子群体的结果差异如何，学校是否公布？
A school that answers all four candidly, including uncomfortable numbers, is providing more useful information than one with a higher headline figure and no methodology. Progression data are ultimately a tool for asking better questions, not for producing a single ranking.
一所愿意坦率回答全部四个问题、包括不体面数字的学校，比一所核心数字更高却毫无方法说明的学校提供的信息更有价值。升学去向数据归根结底是提出更好问题的工具，而不是用来生成单一排名的工具。
Published by TutorHao | Education Studies 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