📚 Why Attendance Data Is Not Published School by School | 出勤数据为什么不逐校公布而按学段汇总
Anyone who has tried to build a national picture of school attendance quickly notices something odd. National figures appear every fortnight, regional and local authority breakdowns follow, and schools publish their own absence percentages on their websites – yet there is no single, clean, downloadable table ranking every school in England by attendance. The reason is not secrecy for its own sake. It is a deliberate design choice built on statistical robustness, data protection law, accountability theory and the practical architecture of how registers are actually collected.
任何试图勾勒全国学校出勤全貌的人都会很快发现一件奇怪的事：全国数据每两周公布一次，地区和地方政府层面的细分数据随之而来，学校也会在自己的网站上公布缺勤率——但始终不存在一份干净、可下载、按出勤率给英格兰每所学校排名的完整表格。原因并非单纯的保密，而是一种刻意的制度设计，建立在统计稳健性、数据保护法、问责理论，以及学籍登记实际采集方式之上。
1. Framing the Question | 先厘清问题的前提
The first thing to establish is that the premise is only partly true. Some school-level absence figures have been published for years. The annual school performance data released by the Department for Education (DfE) has at times included school-level overall absence and persistent absence measures, and every maintained school must publish its own absence figures on its website. What is genuinely withheld is high-frequency, comparable, school-by-school data – the kind that would allow a live league table of attendance.
首先要澄清的是，这个前提只对了一半。学校层面的部分缺勤数据其实已经公布多年。教育部（DfE）发布的年度学校表现数据有时就包含学校层面的总体缺勤率和持续缺勤率，而且每一所公立学校都必须在自己的网站上公布本校缺勤数据。真正被保留的，是高频、可比、逐校的数据——也就是能够形成实时出勤排行榜的那类数据。
So the real question is not ‘why is school-level data hidden?’ but ‘why is the boundary drawn between annual, lagged, context-rich school measures and daily, live, aggregate measures?’ That boundary is the subject of this article.
所以真正的问题不是”为什么学校层面的数据被藏起来”，而是”为什么这条界线画在年度、滞后、带背景信息的学校指标与每日、实时、汇总指标之间？”这条界线正是本文的主题。
2. How Attendance Data Is Actually Collected | 出勤数据究竟是如何采集的
Attendance is recorded in the school register, and the unit of measurement is the session, not the day. A session is half a school day: morning or afternoon. A pupil in a school with 190 teaching days therefore has 380 possible sessions in an academic year. Every session is coded – present, authorised absence, unauthorised absence, or a statistical code such as ‘B’ for off-site educational activity.
出勤记录写在学校学籍册（register）中，计量单位是”节次”（session），而不是”天”。一个节次是半个教学日：上午或下午。因此，在一学年有 190 个教学日的学校里，一名学生全年共有 380 个可能节次。每个节次都会被编码——出席、获准缺勤、未获准缺勤，或诸如”B”（校外教育活动）之类的统计代码。
Since 2024, all state-funded schools in England have been required to share daily register data with the DfE, which now publishes fortnightly attendance statistics during term time. This is a genuinely new capability: for the first time the department can see attendance nationally with a lag of days rather than months.
自 2024 年起，英格兰所有国家资助学校都必须向教育部共享每日学籍数据，教育部如今在学期内每两周发布一次出勤统计。这是一项真正的新能力：教育部首次能够以”天”而非”月”为滞后单位观察全国出勤状况。
The table below summarises what appears at each level of the system.
下表总结了系统各层级所公布的内容。
| Level | What is published | Frequency |
| National | Overall absence rate, persistent absence rate, severe absence | Fortnightly and annually |
| Regional / local authority | Same measures, broken down by area | Fortnightly and annually |
| Phase | Primary, secondary, special, alternative provision | Fortnightly and annually |
| Pupil group | Free school meals, SEND, EAL, ethnicity, gender | Annually |
| Individual school | School’s own website; occasional annual performance measures (often suppressed) | Annually, lagged |
3. Small Numbers and Statistical Fragility | 小样本与统计脆弱性
The most powerful technical reason for aggregation is sample size. The overall absence rate is a simple proportion:
汇总最有力的技术原因是样本量。总体缺勤率是一个简单的比例：
Overall absence rate = (Total sessions missed ÷ Total possible sessions) × 100
For a large secondary school with 1,200 pupils, the denominator is roughly 1,200 × 380 = 456,000 sessions, so random variation washes out. For a small rural primary with 55 pupils, the denominator is only 20,900 sessions – and one pupil absent for a fortnight removes 20 sessions, moving the rate by around 0.1 percentage points on its own. A single family holiday in term time or one long-term illness can shift a small school’s published figure by several percentage points.
对于一所有 1200 名学生的大型中学，分母约为 1200 × 380 = 456000 个节次，随机波动很容易被平均掉。而对于一所只有 55 名学生的乡村小学，分母仅为 20900 个节次——一名学生缺勤两周就会减少 20 个节次，单凭这一项就能让缺勤率变动约 0.1 个百分点。一次学期内家庭旅行或一例长期疾病，就可能让一所小规模学校的公布数字变动好几个百分点。
This matters because parents, journalists and even some policymakers read a percentage as a stable property of a school, when in a small school it may be largely noise. Statistical agencies therefore apply suppression rules: below a minimum cohort or minimum number of sessions, figures are withheld or banded rather than published as exact numbers.
这一点之所以重要，是因为家长、记者甚至部分政策制定者会把一个百分比当作学校”稳定的属性”来解读，而在小规模学校里，它可能主要是噪声。因此统计机构会设置抑制规则：当学生人数或节次总数低于某个下限时，数据会被隐去或以区间形式公布，而不给出精确数字。
4. Data Protection and Pupil Confidentiality | 数据保护与学生隐私
The second reason is legal. Attendance records are personal data about identifiable children, and processing them is governed by the UK GDPR and the Data Protection Act 2018. Publishing school-level figures is not automatically unlawful, but it must satisfy the data minimisation and purpose limitation principles in Article 5, and statistical processing for public interest purposes receives specific safeguards under Article 89.
第二个原因是法律层面的。出勤记录属于涉及可识别儿童的个人数据，其处理受英国《通用数据保护条例》（UK GDPR）和 2018 年《数据保护法》约束。公布学校层面数据本身并不必然违法，但必须满足第 5 条的数据最小化和目的限制原则；而为公共利益进行的统计处理，在第 89 条下享有特定的保障。
The risk is re-identification. In a special school with 30 pupils, or a small sixth form, a breakdown by phase, year group and pupil characteristic can effectively single out one child. Statisticians call the cumulative version of this the mosaic effect: individually harmless fragments combine into an identifiable individual.
风险在于重新识别（re-identification）。在一所只有 30 名学生的特殊学校，或规模很小的六年级部，按学段、年级和学生特征交叉细分，实际上就可能锁定某一个孩子。统计学家把这种累积效应称为”马赛克效应”：单独看无害的碎片信息，组合起来就能识别出具体个人。
That is why disclosure control techniques – suppression thresholds, rounding, banding into categories such as ‘above average’ – are standard practice, and why the smaller the unit of publication, the heavier the protection must be.
这正是为什么披露控制技术——抑制阈值、四舍五入、归入”高于平均水平”之类的区间——属于标准做法，也是为什么公布单位越小，保护措施就必须越严。
5. The League Table Problem | 排行榜问题
Aggregation also reflects a political and ethical judgement about what a published number does once it is released. Attendance is influenced by factors largely outside a school’s control, so a school-by-school ranking would function less as an accountability tool and more as a proxy for neighbourhood poverty.
汇总还体现了一种政治与伦理判断：一个数字一旦公布，会产生什么后果。出勤受许多学校几乎无法控制的因素影响，因此逐校排名与其说是问责工具，不如说会变成社区贫困程度的替身指标。
There is a statistical trap here too. Even where two schools have identical underlying performance, the smaller one will appear more extreme in any given year, so a raw ranking systematically punishes small schools and rewards large ones – an example of the broader phenomenon known as regression to the mean.
这里还有一个统计陷阱。即便两所学校的真实表现完全相同，规模较小的那所在任何一年里都会显得更极端，因此原始排名会系统性地惩罚小学校、奖励大学校——这是”均值回归”这一更广泛现象的一个例子。
Aggregating by phase sidesteps much of this. A phase-level figure tells you about a system, not about a particular head teacher’s competence, and it cannot be used as a shopping guide for house prices.
按学段汇总在很大程度上绕开了这些问题。学段层面的数字讲述的是整个系统的情况，而不是某位校长的胜任程度，也无法被用作房价的购物指南。
6. Context: Deprivation, SEND and Mobility | 背景因素：贫困、特殊教育需求与流动性
Attendance correlates strongly with disadvantage. Pupils eligible for free school meals, pupils with special educational needs and disabilities (SEND), pupils with English as an additional language, and pupils who change school mid-year all show systematically different absence patterns. Emotionally based school avoidance and mental health difficulties are also unevenly distributed.
出勤与弱势处境高度相关。符合免费校餐资格的学生、有特殊教育需求与残障（SEND）的学生、英语为附加语言的学生，以及学年中途转学的学生，其缺勤模式都存在系统性差异。情绪性拒学与心理健康问题同样分布不均。
A school-level table without these context variables invites the wrong comparison. A school in a coastal town with high deprivation and a transient population may record 12 per cent absence while doing excellent pastoral work; a leafy suburban school may record 4 per cent with far less effort.
缺少这些背景变量的逐校表格会诱发错误的比较。一个位于高贫困、人口流动大的海滨城镇的学校，可能在做着出色的学生关怀工作却录得 12% 的缺勤率；而一所绿树成荫的郊区学校可能只付出很少努力就录得 4%。
Phase-level aggregation does not solve this problem entirely – secondary absence is structurally higher than primary absence – but it separates schools facing broadly similar structural conditions, which is a more defensible basis for comparison.
按学段汇总并不能完全解决这个问题——中学缺勤率在结构上本就高于小学——但它把面临大体相似结构性条件的学校归入同一组，这是更具辩护力的比较基础。
7. Perverse Incentives and Gaming | 逆向激励与数据博弈
Whenever a metric becomes high-stakes, behaviour bends towards the metric. Attendance is no exception. Schools have an incentive to manage the register as well as the pupil: choosing codes that move absence from the unauthorised column to the authorised one, recording off-site provision under code ‘B’, or arranging managed moves that transfer a persistently absent pupil to another school’s roll.
任何指标一旦变成高风险指标，行为就会朝指标本身倾斜。出勤也不例外。学校既有动机管理学生，也有动机管理学籍册：选择能把缺勤从未获准一栏挪到获准一栏的代码，用”B”代码记录校外教育安排，或通过”协商转学”把长期缺勤的学生转到另一所学校的名册上。
Publishing live school-level data would sharpen these incentives considerably. It would also create pressure to close registers early, to discourage applications from pupils likely to be absent, and to reclassify difficult cases. Aggregated publication reduces the payoff from such strategies because no single school can move a national or phase-level figure.
公布实时的逐校数据会显著强化这些激励。它还会带来提前关闭学籍登记、劝阻可能有缺勤风险的学生申请入学，以及对棘手个案重新分类的压力。汇总公布则降低了这类策略的收益，因为没有哪一所学校能左右全国或学段层面的数字。
This is the classic distinction between accountability for outcomes and accountability for processes: the system wants schools to improve attendance, not to improve the appearance of attendance.
这正是”结果问责”与”过程问责”的经典区别：系统希望学校真正改善出勤，而不是改善出勤的”外表”。
8. Who Actually Sees School-Level Data | 谁真的能看到学校层面的数据
The claim that school-level attendance data is not published does not mean nobody sees it. The DfE holds daily data for every school. Ofsted inspectors receive an Inspection Data Summary Report containing the school’s attendance profile alongside local and national comparators. Local authority attendance teams, multi-academy trust leaders and Regional Improvement for Standards and Excellence teams work with named school data constantly.
说学校层面的出勤数据不公布，并不意味着没人能看到它。教育部掌握着每所学校的每日数据。教育标准局（Ofsted）的督学在视察前会收到”视察数据摘要报告”，其中包含该校的出勤概况及地方与全国对照数据。地方政府出勤团队、多学院信托领导层以及区域标准与卓越改进团队，日常都在使用具名的学校数据。
Parents have a direct route too: schools must publish their own absence figures, and any parent can ask the school or governing body for the underlying numbers.
家长也有直接途径：学校必须公布本校缺勤数据，任何家长都可以向学校或校董会索取底层数字。
The design principle is therefore not opacity but graduated access: the more identifiable and the more granular the data, the tighter the circle of legitimate users; the more aggregated, the wider the publication.
因此设计原则并非不透明，而是分层访问：数据越具可识别性、越细颗粒度，合法使用者的圈子就越小；数据越汇总，公布范围就越广。
9. Definitions and Thresholds | 定义与阈值
Much public confusion about attendance comes from definitions rather than from suppression. The categories below are the ones that matter for any exam answer or policy discussion.
公众对出勤的困惑，很多来自定义而非数据抑制。以下类别是任何考试作答或政策讨论中都必须掌握的。
| Term | Definition |
| Session | Half a school day; 2 sessions per day, 380 per year (190 days) |
| Authorised absence | Absence with school permission (illness, medical appointment, bereavement, religious observance) |
| Unauthorised absence | Absence without permission, including unauthorised term-time holiday |
| Persistent absence | Missing 10% or more of sessions, i.e. attendance at or below 90% |
| Severe absence | Missing 50% or more of sessions, i.e. attendance at or below 50% |
Note the asymmetry these thresholds create. A pupil who misses every Friday afternoon for a year sits just under the persistent absence line; a pupil who misses a month through illness may cross it. This is why phase-level rates are published alongside pupil-group breakdowns – the headline number alone cannot distinguish between very different underlying stories.
注意这些阈值造成的不对称性。一名学生若整年每周五下午缺勤，会刚好卡在持续缺勤线下方；另一名学生因病缺勤一个月，却可能越过这条线。这就是为什么学段层面的比率要和学生群体细分数据同时公布——单看总数字，无法区分截然不同的底层情形。
10. Why Phase Is the Policy Sweet Spot | 为什么按学段汇总是”政策最优颗粒度”
Phase – primary, secondary, special, alternative provision – turns out to be the level at which several requirements intersect at once. It is large enough for statistical stability, since each phase contains hundreds of thousands of pupils. It is stable in composition year on year, so trends are meaningful. And it maps directly onto how policy levers differ.
学段
Published by TutorHao | Exam Prep 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