📚 Counting Home-Educated Students: Official Statistical Definitions Explained | 在家教育学生人数的官方统计口径
Ask a simple question - how many children are educated at home? - and you will get answers ranging from about one per cent of the school-age population to more than five per cent, depending on which country, which agency and which definition you use. That gap is not a measurement error in the ordinary sense; it is a definitional gap. Home education is counted through administrative registers in some jurisdictions, through household surveys in others, and through modelling and advocacy estimates in the rest. This article unpacks the official statistical definitions behind the headline figures, the sampling and non-sampling errors that surround them, and the ways in which A-Level Sociology and Mathematics students can use these ideas in examination answers.
如果问一个简单的问题——有多少儿童在家接受教育？——你会得到从占学龄人口约 1% 到超过 5% 不等的答案，取决于你采用哪个国家、哪个机构、哪一套定义。这一差距并非普通意义上的测量误差，而是定义口径的差距。有的辖区通过行政登记来统计在家教育，有的通过家庭抽样调查，其余的则依靠建模和倡议团体的估算。本文拆解这些头条数字背后的官方统计口径、围绕它们的抽样误差与非抽样误差，以及 A-Level 社会学与数学（统计）考生可以如何在答题中运用这些概念。
1. Why the Count Is Contested | 为什么这一数字充满争议
Home education sits at the intersection of three competing policy interests: parental autonomy, child safeguarding and public expenditure. The number is therefore never merely descriptive. A low figure supports the argument that home education is a marginal practice requiring no new regulation; a high figure supports claims about safeguarding blind spots, funding shifts and the need for mandatory registers. Statisticians describe this as the political salience of a statistic: the higher the salience, the more pressure there is on the definition itself.
在家教育处于三重政策利益的交汇点：家长自主权、儿童保护与公共财政支出。因此这个数字从来不只是描述性的。数字偏低，可以支持"在家教育只是边缘现象、无需新监管"的论点；数字偏高，则可以支持关于保护盲区、经费转移以及建立强制登记制度必要性的主张。统计学家称之为统计数字的政治显著性：显著性越高，定义本身承受的压力就越大。
For A-Level purposes, the key insight is the one made by interpretivist sociologists: official statistics are socially constructed. They are not neutral reflections of reality but products of decisions about who to count, when to count them and who is permitted to report them. The same substantive phenomenon can yield figures that differ by a factor of three without anyone making a factual error.
就 A-Level 考点而言，关键洞见正是诠释派社会学家的主张：官方统计是社会建构的产物。它们并非对现实的neutral反映，而是关于"统计谁、何时统计、由谁报告"等一连串决定的产物。同一种社会现象可以得出相差三倍的数字，而其中没有任何一方犯下事实性错误。
2. Core Definitions: EHE, EOTAS and Homeschooling | 核心定义：选修式在家教育、校外教育与在家上学
The single biggest driver of variation is the definition of the counted population. Four terms recur in official documents and they are not interchangeable.
造成差异的最大单一因素，是被统计人口的定义。官方文件中反复出现四个术语，它们并不能互换使用。
- Elective home education (EHE) - the parents have chosen, without pressure from the school or local authority, to educate the child at home rather than at school. This is the narrowest and most commonly used definition in England.
- 选修式在家教育（EHE）——父母在没有学校或地方当局施压的情况下，主动选择让孩子在家而非在校接受教育。这是英格兰最窄、也最常用的定义。
- Education otherwise than at school (EOTAS) - education arranged by the local authority because school is not suitable, for example for medical or behavioural reasons. These children are often in home or alternative provision but were not chosen by the parent.
- 校外教育（EOTAS）——由地方当局安排的、因学校不适合（如医疗或行为原因）而在校外进行的教育。这些儿童常在家庭或替代性教育机构中，但并非家长主动选择。
- Homeschooling (US usage) - a parent-reported arrangement in which the child is taught at home instead of attending a regular school for at least part of the school year.
- 在家上学（美国用法）——家长自报的安排：孩子在学年中至少有一部分时间在家接受教学，而非进入常规学校。
- Children missing education (CME) - children of compulsory school age who are not registered at any school and not receiving suitable education elsewhere. This is a safeguarding category, not a parental-choice category.
- 失学儿童（CME）——处于义务教育年龄、未在任何学校注册且未在其他地方接受适当教育的儿童。这是一个保护性分类，而非家长选择分类。
A crucial boundary case is remote schooling. During the pandemic many children were taught at home by a school teacher over a video link. Official definitions in the United States and the United Kingdom treat that as school enrolment, not as home education, because the instruction is provided and directed by a school. Getting this boundary wrong inflates the home education figure substantially.
一个关键的边界情形是远程上学。疫情期间，许多儿童在家通过视频由学校教师授课。美国与英国的官方定义将其视为学校在册学生，而非在家教育，因为教学由学校提供并主导。混淆这一边界会显著夸大在家教育的数字。
3. Administrative Data versus Household Surveys | 行政数据与家庭抽样调查
There are two main data-generating processes for home education counts, and they answer slightly different questions.
在家教育人数的统计主要有两种数据生成过程，它们回答的问题略有不同。
| Dimension | Administrative register | Household survey |
| Who is counted | Only children notified to the authority | All children in sampled households, whether notified or not |
| Unit | Individual child, usually by term or census date | Child within household, weighted to population |
| Main error type | Coverage error (undercount of non-notified families) | Sampling error and response bias |
| Timing | Point-in-time snapshot | Reference period, often "during this school year" |
| Best used for | Local service planning, safeguarding | National prevalence and trend estimation |
The two sources rarely agree. A register measures the administrative footprint of home education; a survey measures its social prevalence. When a country has both, the survey figure is normally higher because it captures families who never told anyone.
两种数据来源很少一致。登记数据衡量的是在家教育的行政足迹；抽样调查衡量的是它的社会流行程度。当一国同时拥有两者时，调查数字通常更高，因为它能捕捉到从未向任何机构通报的家庭。
4. The US Model: The NHES Parent-Reported Definition | 美国模式：NHES 的家长自报口径
The United States produces the most widely cited national estimate through the National Household Education Surveys Program (NHES), administered by the National Center for Education Statistics (NCES). The relevant instrument is the Parent and Family Involvement in Education survey, which screens a large national sample of households and then asks detailed questions of parents with school-age children.
美国通过国家教育统计中心（NCES）主持的"全国家庭教育调查项目"（NHES）发布最被广泛引用的全国估计。相关工具是"家长与家庭参与教育调查"，它先筛查一个规模庞大的全国家庭样本，再对有学龄子女的家长提出详细问题。
The official definition has three elements, and each one excludes some families that a layperson would call homeschooled:
官方定义包含三个要素，每一个都排除了一些普通公众会称为"在家上学"的家庭：
- The child is of school age and in the relevant grade range (kindergarten through grade 12).
- 孩子处于学龄阶段且属于相应年级段（幼儿园至 12 年级）。
- The child is taught at home instead of at a regular school for at least part of the school year.
- 孩子在学年中至少有一部分时间在家接受教学，而非进入常规学校。
- Children enrolled in a public or private school who receive supplementary home instruction are excluded, as are children enrolled in virtual or online public schools, who count as public school students.
- 已在公立或私立学校注册、仅额外接受家庭辅导的儿童被排除；注册在虚拟或在线公立学校的儿童也被排除，他们计入公立学校学生。
Because the estimate comes from a sample, it is published with a standard error and a confidence interval. NCES typically reports that the percentage of students homeschooled has fluctuated across recent waves, with point estimates in the low single digits in ordinary years and markedly higher readings during the pandemic period, when many families temporarily moved to home-based instruction. Analysts must therefore distinguish a structural trend from a pandemic shock, which is only possible by comparing multiple waves.
由于估计值来自样本，发布会同时给出标准误与置信区间。NCES 通常报告：近年来在家上学学生比例的估计值在不同调查轮次间波动，正常年份的点估计处于个位数低位，而疫情期间读数明显上升，因为许多家庭暂时转向居家教学。因此分析者必须区分结构性趋势与疫情冲击，而这只能通过比较多个调查轮次来实现。
5. The UK Model: Local Authority Returns and DfE Collections | 英国模式：地方当局回报与教育部数据
England has no single historic register of home-educated children. Until recently the Department for Education relied on a non-statutory annual collection from local authorities, which asked each authority how many children in its area were known to be in elective home education. Coverage was uneven: a well-resourced authority with an active EHE team reported more children than a neighbouring authority that did not chase notifications.
英格兰历史上没有统一的在家教育儿童登记册。直到最近，教育部依赖地方当局的自愿性年度回报，询问各辖区有多少儿童被知晓处于选修式在家教育中。覆盖情况并不均衡：设有专职 EHE 团队、资源充足的辖区，回报的儿童数会高于不去追查通报的邻近辖区。
The data collection has since been formalised and moved to a more regular cycle, and policy proposals for statutory registers of children not in school have been debated in Parliament. Published DfE figures have shown a clear upward movement: authorities reported roughly 92,000 children in elective home education in one recent year and around 111,000 in the following year, although exact totals are revised between releases and readers should always check the current statistical bulletin.
此后数据收集被正式化并转入更规律的周期，关于建立"失学儿童法定登记册"的政策提案也已在议会中讨论。教育部已发布的数字显示出明确上升趋势：地方当局在某一年回报约 92,000 名儿童处于选修式在家教育，下一年约 111,000 名；不过具体总数会在各次发布之间修订，读者应始终查阅最新的统计公报。
Two caveats matter for interpretation. First, the figure counts children known to authorities, so it is a floor rather than a ceiling. Second, it covers EHE only. Independent work by the Children's Commissioner for England and by the Association of Directors of Children's Services has put the wider population of children not in school - including those awaiting a place, those in unregistered provision and those missing education entirely - at a considerably higher level, sometimes cited at around 150,000 at any one time.
两点解释上的提醒很重要。第一，该数字统计的是"被当局知晓"的儿童，因此它是下限而非上限。第二，它只覆盖 EHE。英格兰儿童事务专员以及儿童服务主任协会的独立研究显示，更广义的失学儿童群体（包括等待学位、处于未注册教育机构、以及完全失学的儿童）规模要大得多，常被引用为任一时刻约 15 万人。
Attendance statistics provide a further definitional trap. Persistent absence is defined as missing 10 per cent or more of possible school sessions, and severe absence as missing 50 per cent or more. These are measures of enrolled pupils, not of home-educated children, and conflating them is a common examination error.
出勤统计还提供了另一个定义陷阱。"持续性缺勤"定义为缺课达到可能课节的 10% 或以上，"严重缺勤"为 50% 或以上。这些指标衡量的是在册学生，而不是在家教育儿童，把两者混为一谈是常见的答题错误。
6. Comparators: Ireland, Australia and Canada | 其他辖区：爱尔兰、澳大利亚与加拿大
Other English-speaking jurisdictions illustrate how the legal basis shapes the count. Ireland operates a statutory notification route: parents must apply to register with the Child and Family Agency, and the agency assesses whether the education provided meets constitutional requirements. Because registration is a legal duty, the Irish count is closer to a true register, but it is still limited to families who comply.
其他英语辖区可以说明法律基础如何塑造统计数字。爱尔兰设有法定通报渠道：家长必须向儿童与家庭事务署申请登记，由该署评估所提供的教育是否符合宪法要求。由于登记是法定义务，爱尔兰的数字更接近真正的登记册，但仍仅限于遵守规定的家庭。
Australia regulates home education at state and territory level, so there is no single national register; national totals are built up from jurisdictional counts and supplemented by survey and research estimates. Canada is similar, with provincial oversight and considerable variation in whether notification is required, encouraged or effectively unenforced.
澳大利亚在州与领地层面监管在家教育，因此没有统一的全国登记册；全国总数由各辖区数字汇总，并以调查与研究估计作补充。加拿大情况类似，由各省监管，在通报是"必需""鼓励"还是"事实上未执行"之间存在相当大的差异。
At supranational level, the OECD and Eurostat collect enrolment-based data mapped to ISCED levels. Because these collections count children in registered educational institutions, home-educated children are largely invisible in comparative tables such as Education at a Glance. Cross-national datasets therefore tend to understate home education systematically rather than randomly.
在超国家层面，经合组织与欧盟统计局收集以在册学生为基础、映射到 ISCED 层级的数据。由于这些收集统计的是在注册教育机构中的儿童，在家教育儿童在《教育概览》之类的比较表中基本不可见。因此跨国数据往往系统性地、而非随机地低估在家教育。
7. Six Variables That Move the Number | 使数字变动的六个变量
When you compare two home education figures, check these six variables before concluding that practice has changed.
当你比较两个在家教育数字时，在下结论说"做法发生了变化"之前，请先检查以下六个变量。
- Who counts as home educated - pure EHE, EHE plus EOTAS, or all children not in school.
- 谁算作在家教育——纯 EHE、EHE 加 EOTAS，还是所有失学儿童。
- Part-time or full-time - a child who attends school three days a week and is taught at home for two days is counted differently in different systems.
- 部分时间还是全部时间——一周三天在校、两天在家受教的儿童，在不同体系中的归类不同。
- Point-in-time or cumulative - a snapshot on a census date versus "at any point during the year". Cumulative counts are always larger.
- 时点还是累计——普查日当天的快照，对比"年内任一时刻"。累计口径总是更大。
- Unit of measurement - children or families. A family with four home-educated children is one household but four children, so a rate per 1,000 families cannot be compared with a rate per 1,000 children.
- 计量单位——儿童还是家庭。一个有四名在家教育子女的家庭是一个住户，但是四名儿童，因此"每千家庭比率"不能与"每千儿童比率"直接比较。
- Age and grade range - compulsory school age only, or including pre-school and post-16 learners.
- 年龄与年级范围——仅义务教育年龄，还是包含学前与 16 岁后学习者。
- Who reports - a local authority officer, a parent in a household survey, or an advocacy organisation estimating from its own membership.
- 由谁报告——地方当局官员、家庭调查中的家长，还是依据自身成员数据估算的倡议组织。
In an examination answer, showing awareness that a change in the reported figure may reflect a change in definition rather than a change in behaviour is a mark of high-level analysis.
在考试作答中，能指出"报告数字的变化可能源于定义变化而非行为变化"，是高阶分析的标志。
8. Sampling Error and Confidence Intervals | 抽样误差与置信区间
Survey-based estimates of home education are proportions, so their precision can be quantified. The estimated proportion is simply the weighted count of homeschooled children divided by the weighted count of all school-age children in the sample.
基于调查的在家教育估计值是比例，因此其精确度可以被量化。估计比例就是样本中在家教育儿童的加权计数除以全部学龄儿童的加权计数。
p̂ = Σ wᵢ xᵢ ÷ Σ wᵢ
For a simple random sample the standard error of that proportion is calculated as follows, and the margin of error at 95 per cent confidence is approximately 1.96 times the standard error.
对于
Published by TutorHao | Exam Prep 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