Past Papers开刷时机 三遍法 错题归因

Past Papers开刷时机 三遍法 错题归因

每年十月开始，问真题什么时候刷的学生就多起来。这个问题没有统一答案，但有一个可以检查的节点：把考试局 spec 上的知识点过完六成，再动整卷。太早做整卷，错的地方全是还没学过的内容，错题本记满却学不到东西。太晚做，第一次计时就会撞上时间压力，连会做的题都写不完。下面按起点、做法、判分、排期四件事拆开讲。

Every October the same question arrives: when should past papers start. There is no single date that fits everyone, yet one checkable threshold works across AQA, Edexcel and OCR specifications, namely roughly six in ten topics taught before a full timed paper is worth sitting; earlier than that, the errors recorded are gaps in teaching rather than exam mistakes, and later the first timed attempt collides with real time pressure. What follows separates the timing question from the method question.

第一套整卷的起点：大纲覆盖六成

判断能不能开刷整卷，看的是 spec 而不是月份。AQA、Edexcel、OCR 的 spec 都把内容按 topic 列清，先数一遍自己学过哪些。超过六成才适合做整卷，剩下四成没学完时，计时结果没有参考价值。学校进度慢的话，先做按 topic 拆开的旧题，只做学过的那几道。整卷真正考的是时间分配和跨 topic 切换，这两项在零散做题里练不出来。很多学生以为真题越早开始越好，实际上开得太早只是把不会的题抄了一遍。第一次整卷不用挑最新的，挑三年前的那套更合适，因为它的公开版一定拿得到。

Whether a full paper measures exam skill or simply exposes missing teaching depends on specification coverage, and counting taught topics against the AQA or Edexcel specification gives a rough but workable answer; below about six in ten, topic-split questions from older papers do more useful work, because time allocation and switching between topics cannot be practised on isolated questions. Many students assume earlier is always better; the notebooks kept in September suggest otherwise. For a first timed paper an older series is the safer choice, since its public version is guaranteed to be available.

真题三遍法：计时、归因、只重错题

同一套卷子做三遍才划算，但三遍做的事完全不同。第一遍严格按正式时长计时，早上九点或下午一点半开始，中间不翻书不查手机。第二遍不看分数，把每道错题归到知识缺口、审题偏差、时间压力三类里。第三遍只做归出来的错题，隔两周重做一次，做对才算过。常见的错误是第二遍直接看答案，看懂就翻过去，三周后同样的题还是错。第三遍的时间成本很低，一套卷子通常四十分钟就够。做完三遍的卷子不要收进柜子，把它当错题索引，考前翻目录比翻整本笔记快。

Three passes over a single paper serve different purposes; the first is a timed rehearsal under real conditions, the second is diagnosis rather than scoring, and the third is a repeat of the diagnosed items a fortnight later. Checking answers straight after a paper feels productive, yet recognising a solution is not the same as producing it, which explains why the same question fails again three weeks on. The repeat pass is cheap, usually under an hour for a whole paper, and keeping the script as an index of errors beats filing it away.

mark scheme 按点给分：把答案切成得分点

对答案的方式决定真题的价值。AQA 和 Edexcel 的 mark scheme 都是按点给分，每一分对应一个可辨认的表述，写对了才给分，写得多不给额外的分。对答案时把自己答案的每一句和 mark scheme 的每一条并排放，标出哪一句命中了哪一分。没命中的部分才是丢分的真实位置，只标对错学不到东西。理科的计算题还要看 method mark，过程对了结果错通常仍有一到两分。mark scheme 的措辞往往比教材更严格，术语差一个词就不给分。阅卷时最常看到的是答案对但过程跳步，method mark 连同后面的 accuracy mark 一起丢掉。

Mark schemes award marks point by point, and each mark sits on a recognisable statement; laying every sentence of a script beside the scheme shows exactly which statements earned marks, whereas marking whole answers right or wrong teaches very little. Method marks survive a wrong final value in most science papers provided the working is visible, and scheme wording is stricter than textbook prose, so a near-miss term rarely earns anything. What examiners see most often is a correct answer with skipped steps, which loses the method mark together with the accuracy mark that depends on it.

examiner report 的高频失分点怎么读

每个考季结束后，AQA、Edexcel、OCR 都会发布 reports on the exam，逐题写学生错在哪里。AQA 的官方页面写明，reports on the exam 在放榜前一天上线，学校的 Centre Services 端口更早。这份报告列的是当年真实卷面上的错误，不是编出来的模拟题，比任何复习资料都直接。读法是先看自己错的那几道题对应的题号，再看整份报告里反复出现的表述问题。写作类科目里，报告经常点出答非所问和缺少文本证据这两条，它们每年重复。很多学生从头读到尾，读完只记住几句评价，等于没读。边读边在自己的错题上标注对应条目，报告才真正用上。

Reports on the exam appear after every series and describe, question by question, where candidates lost marks; the AQA website says the report goes live the day before results day, with centre access earlier. Reading them from cover to cover produces a vague impression, while reading the sections that match your own wrong questions produces a checklist, especially in essay subjects where the same two criticisms recur each year, answering a different question and quoting without textual evidence. Annotating those lines onto your own scripts is what turns the report into something usable.

错题归因三类：知识缺口、审题偏差、时间压力

归因错了，后面的练习全是白做。知识缺口的表现是看答案也看不懂，这类要回教材和 spec 重学。审题偏差的表现是看完答案立刻会做，题干里的限定词被忽略了，这类要练拆题干。时间压力的表现是单独做能做对，整卷里空着，这类要练切分时间而不是补知识。三类错题的处理方式完全不同，混在一起记进错题本，等于把三种问题当一种治。错得最多的是审题偏差，因为它看起来像粗心，实际是没读懂指令词。错题本上每道题后面写清类别，比抄题目本身有用得多。

Misdiagnosis wastes every hour of practice that follows; a gap in knowledge shows up when the worked solution is still incomprehensible, whereas a reading error disappears the moment the answer is seen, and genuine time pressure only appears when the clock runs. Treating these three as one problem in a single notebook is why revision stalls. Reading errors are the most common of the three, largely because they look like carelessness rather than a misread command word, and labelling each entry by type in the margin does more than copying the question out.

卷面时间压缩一成：把时间压力提前

正式考试的时长是给写得完的人设的，练习时需要更紧。做法是把卷面时间压到正式时长的九成，两小时的卷子按一百零八分钟做，逼自己在压力下取舍。压缩训练的目标不是做完全卷，而是学会先拿容易的分，把卡住的题标号跳过。回到正常时长时，会明显感觉到余量，这种余量在考场上就是检查的时间。这一步最容易做错的是只压缩时间不改策略，结果每次都在最后两题上崩掉。计时用手机秒表，别用教室挂钟，五秒的误差在整卷里会放大。每次记录卡住的时间点，三次之后规律会自己浮出来。

Practice papers should run tighter than the real thing, so cutting the clock to ninety per cent of the published duration forces decisions about what to skip; the aim is not to finish but to bank easy marks first and flag the questions that stall. Once the clock returns to its real length the spare minutes feel obvious, and that surplus is what pays for checking. Compressing time without changing strategy simply reproduces the same collapse in the last two questions, and a stopwatch on the desk reveals which questions are eating the minutes.

最近一套卷拿不到：官网要等一年

学生常常找不到当年最新的一套真题，这不是资料缺失。AQA 官网写明，试卷在考试季结束一年后才放到公开网站，2026 年夏季的卷子要到 2027 年 7 月 9 日才上线。公开网站保留三年，学校的 Centre Services 保留四年，学校端口在考试后十天就能拿到题。所以学生手里最新的一套公开卷，通常比当年考试晚一年，这属于正常情况。Edexcel 和 OCR 的做法类似，最新卷优先给学校做模考。急着要最新卷，可以让老师从校内端口打印一份，不要去买来路不明的 PDF。用晚一年的卷子备考，完全不影响效果。

Not finding the most recent paper is normal rather than a gap in the market; AQA publishes question papers on its public website one year after a series, which puts the summer 2026 papers online on 9 July 2027, while centres see them about ten days after each exam through Centre Services. Edexcel and OCR follow a similar pattern, keeping the newest scripts for school mocks. A paper a year old is perfectly good preparation, and a school printout is a safer route than an unattributed PDF.

考前十二周排期：从考试周往前倒推

排期从考试周往前倒推，比从今天往后数有效。A-level 和 GCSE 都在五月中下旬开考，按 JCQ 统一的时段，上午九点、下午一点半各一场。以第一场考试为基准往前十二周，正好落在二月下旬，这段时间要完成六到八套完整计时卷。考前四周开始只做错题和整卷复盘，不再开新卷。学校模考周会打断节奏，把排期写进日历比记在脑子里可靠。每周固定一个时段做整卷，比等到有空再做实际得多。模考成绩不要当成预估分，它只是排期里的一个检查点。

Working backwards from the first exam works better than counting forwards from today; with A-level and GCSE papers starting in mid to late May under the JCQ common timetable of nine in the morning and half past one, twelve weeks before the first paper lands in late February, which is when six to eight full timed papers should be completed. The last four weeks belong to error review rather than new papers, and a fixed weekly slot survives the disruption of school mock weeks better than good intentions. Mock results are a checkpoint in that schedule, not a predicted grade.

分数线每年重划：别用旧分数线估分

分数线每年按当季卷面的实际难度重划，不是固定数值。AQA 和 Edexcel 都在放榜日公布当季的 grade boundaries，同一科目不同年份能差好几分。用三年前的分数线换算自己现在的分数，结论往往是错的。更有用的做法是记录 raw mark 和当季分数线之间的差值，看自己离上一档差几分。这个差值在几次整卷之后会稳定下来，比百分比可靠。看分数线时注意区分 total 和 component，两个层级的线不一样。很多学生只看总分线，忽略了自己最弱的那张卷子拉了多少。

Grade boundaries are redrawn for each series according to how the papers performed, so a fixed number carried over from an old year will mislead; AQA and Edexcel publish boundaries on results day, and the same subject can move by several marks between years. Tracking the distance between your raw mark and the boundary above you is far more useful than tracking a percentage, and that gap settles after a handful of timed papers. Component boundaries matter as much as the overall line, because one weak paper can absorb the whole margin.

刷完全部真题分数不动的原因

把近十年的卷子全部刷完，分数却停在原地，原因通常不在数量。刷题时每套都从头做到尾，做错的题只对答案不改做法，下一次遇到同型题仍按老思路走。真题的价值在于暴露思维习惯，而思维习惯只有在重做时才改得掉。另一个原因是只做同一年份的卷子，题型分布集中，遇到换题型的年份就失效。按 topic 横着做近几年所有卷子的同类题，比按年份竖着刷更容易看出规律。整卷训练和横向训练要交替，只做一种都会留下盲区。最后一条最容易被忽略：做完不记录用时，就永远不知道自己在哪里慢。

Finishing ten years of papers while the score stays flat usually points to method rather than volume; working each paper end to end and only checking answers leaves the underlying approach untouched, so the same question type fails again in the same way. Past papers expose habits, and habits change only on a repeat attempt. Working one topic across several years shows patterns that a single year hides, so horizontal and full-paper practice need to alternate, and timing each attempt is the only way to find out where the minutes actually go.

牛津高阶英汉双解词典 第10版 商务印书馆¥168📱 去京东购买京东精选

更多咨询请联系16621398022（同微信）

需要一对一辅导？Need One-on-One Tutoring?

A-Level / GCSE / IB / KS2-KS3 全科辅导

16621398022 同微信

Visit aleveler.com