IB EE研究问题 选题收窄 评估标准A 4000词
扩展论文是 IB 文凭课程核心的一篇独立研究论文，正文上限 4000 词，学生要在导师指导下自己提出一个可以回答的研究问题。IBO 在 2025 年 2 月出版的《Extended essay guide》把这份论文拆成五条评估标准、共 30 分，A 标准的第一条打分项就是研究问题本身。问题写歪了，后面四千词都在替一句不清楚的话做注脚。这篇文章按选题、收窄、匹配方法、写进引言的顺序，把研究问题拆成可以逐条检查的动作，并把 2027 年首次评估的新标准与旧标准放在一起对照。
The extended essay sits at the centre of the IB Diploma Programme core, capped at 4,000 words and assessed externally, and the research question a student settles in the first weeks quietly decides everything that follows. IBO published a new guide in February 2025 for first assessment in 2027, and under that rubric the research question forms the first strand of Criterion A out of a 30-mark total. In practice, the essays that stall in the final month stalled because the question was too wide to answer in four thousand words, not because the student ran out of sources.
A 标准的第一条打分项就是研究问题
IBO 的评分表把 A 标准叫 Framework for the essay，满分 6 分，第一条打分项就是研究问题。评分描述分三档，最低一档是问题写出来了但与题目不相关、不清楚或不聚焦；中间一档是相关、清楚，但相对全文范围只部分聚焦；最高一档要求相关、清楚，并且在全文范围内聚焦。很多学生以为研究问题只是引言里的一句话，真正被打分的是它与研究方法、结构是否咬合。同一个问题换成另一种研究方法，A 标准的第二条打分项就会掉档。阅卷时看到的失分往往不在文笔，而在这句话能不能撑起四千词的论证。问题模糊时，学生只能在正文里不断换角度，段落之间接不上，最后靠小标题硬撑结构。
Criterion A is called Framework for the essay and carries 6 marks, with the research question as its first strand. The descriptor moves from a question that is stated but unclear, unfocused or unrelated, through one that is clear yet only partly focused, to one that is relevant, clear and focused on the scope of the essay; the middle band is where most competent drafts land. A question that quietly changes its method halfway through drags the second strand down with it, and examiners see that far more often than they see weak prose.
一句能用的研究问题要过四道条件
指南给研究问题列了几条硬条件，要清楚陈述、要聚焦、要能在一个 4000 词的范围里答完，还要能通过系统的研究去检验。把这几个条件翻成学生能用的动作，就是四道自检：这句问句里有没有一个明确的对象、变量或文本；答案是不是需要证据，而不是靠常识就能说完；边界能不能用一两句话写清；四千词的篇幅够不够展开论证。四道里有一道过不去，问题就该往回收。边界写不清时，正文会出现两套互相打架的论点，读者读到第三节还不知道你到底在证明什么。很多学生以为问题写得大一点显得有野心，结果正文只能铺陈描述，评分表里分析那一栏先掉。把问题缩小不是降低难度，是让论证有落点。
A workable question has to be stated clearly, stay focused, be answerable within the 4,000-word limit and lend itself to systematic investigation. Four quick checks cover most of that: is there a named object, variable or text in the sentence, does the answer need evidence rather than common sense, can the boundary be written in two lines, and will four thousand words actually be enough to argue it. Shrinking a question is not a retreat from ambition, since a narrower question is the only kind that can carry a sustained argument.
从宽泛话题收窄成问句的减法
宽泛话题收窄的过程更像减法，删掉的每一层都要能在句子里看出来。写文学的学生常把话题停在莎士比亚很重要这一层，改成问某个角色在某一幕到某一幕之间语言的变化，以及这种变化让读者看到什么，问题立刻可以落到台词上。写科学的学生把咖啡因影响植物生长改成不同浓度的咖啡因溶液对水芹幼苗根长在十四天内的作用，自变量、因变量与观察周期就都齐了。收窄的动作有三个方向：限定对象、限定时间或样本、限定要看的关系。三个方向里选一个先切，句子就短了，论证反而长出来。常见错误是三个方向同时切，最后问题窄到没有资料可查，学生又退回原来那个大题目。
Narrowing usually works as subtraction, and each cut should be visible in the wording of the final question. A literature candidate who starts from Shakespeare ends up asking how one character's language shifts between two specific acts and what that shift lets a reader see, which is a question about lines on a page rather than a reputation. A science candidate turns caffeine and plant growth into a question about three caffeine concentrations and cress root length over fourteen days, at which point the variables and the observation window are already settled; cutting in one direction first keeps the question researchable.
研究问题与研究方法的匹配
A 标准的第二条打分项问的是研究方法与问题是否相配，这一条比很多人想的更硬。问题问的是文本里某种呈现方式，方法就该是细读与对照，而不是问卷；问题问的是某个政策在学生身上的效果，方法就该有样本、有比较组、有可重复的测量。方法写不出数据从哪里来，问题再好也停在设想。很多学生以为方法一节只要写清楚步骤，其实评委看的是这些步骤能不能回答那句问句。方法换了之后，原来的结论段往往整段作废，这也是为什么研究问题要在动笔写正文之前定下来。方法匹配不上时，最省事的补救是把问题往方法能覆盖的范围收一收。
The second strand of Criterion A asks whether the research methods suit the question, and that is a tougher test than a tidy methods section suggests. A question about how something is represented in a text calls for close reading and comparison rather than a questionnaire, while a question about the effect of a policy on students needs a sample, a comparison and a repeatable measurement. Methods that cannot say where the data comes from leave even a good question at the level of a proposal, so the question is best fixed before the body is drafted.
研究问题决定论文骨架与段落分工
A 标准的第三条打分项看结构惯例，结构合不合理要回到问题上去判断。问题里有几个可辨的部分，正文就该有几个对应的论证段，一段负责一个部分，段落之间靠问题里的关系词衔接。问题写成比较两种解释，正文却按时间顺序讲一遍历史，结构分就丢了，因为读者找不到比较。问题写成在多大程度上，正文就要给出程度上的判断与边界，而不是只列正反两面。写提纲时把每个小标题翻译成一句能回答问题的陈述，翻译不出来的小标题通常可以删掉。很多学生以为结构只是排版格式，实际上结构是研究问题在纸面上的展开方式。
Structural conventions form the third strand of Criterion A, and the test is whether the body matches the shape of the question. If the question compares two explanations, the essay needs sections that each carry one side and then weigh them, whereas a chronological account leaves the comparison undone no matter how well written. Drafting an outline in which every heading becomes a sentence that answers part of the question is the quickest way to see which sections are padding.
学科路径里的研究问题怎么写
学科路径要求研究问题落在一门 DP 学科内部，用这门学科的方法与术语来回答。科学类的问句里通常要有自变量、因变量与控制条件，答案要能用测量结果支持，做不到测量的题目会写成文献综述。语言与文学的问句要指向具体文本，比较两首诗的处理方式比讨论某个作家的风格更可行。经济类的问句要能用模型或数据说话，把在多大程度上写成好不好，就跳到价值判断上去了。历史类的问句需要可查的史料范围，问题里最好带时间与地区，否则材料会无限扩张。每门学科的指南都有 subject-specific guidance，动笔前值得把那一节读完再定题。
Within the subject-focused pathway the question has to sit inside one DP subject and be answered with that subject's methods and vocabulary. Science questions usually name an independent variable, a dependent variable and the conditions held constant, because a question that cannot be measured turns into a literature review; language and literature questions point at specific texts, and economics questions need a model or a dataset rather than a verdict on whether something is good. Reading the subject-specific guidance in the IBO guide before fixing the wording saves a term of work.
跨学科路径的研究问题要体现整合
跨学科路径用两门 DP 学科的视角回答同一个问题，研究问题必须真的需要两边一起用，只用一边也能答的题目会被判成整合不足。可行的问句通常把一边的理论与另一边的材料放在同一个句子里，比如用物理的力学去解释某种运动技术中的旋转，或用经济学的激励结构去解释一项健康政策的落地差异。环境系统与社会、文学与表演这两门本身是跨学科科目，只能在学科路径里选，不能放进跨学科路径。整合是在研究过程中长出来的，不必在第一天就做到，但问题里要留下两边交锋的位置。很多学生以为跨学科就是把两个话题拼在一起，评分表看的却是两边是否共同推进了同一个回答。
The interdisciplinary pathway combines two DP subjects, and the question has to need both lenses rather than merely mention them. A workable version puts a theory from one subject against material from the other inside a single sentence, so that neither half could answer it alone; environmental systems and societies and literature and performance are themselves cross-disciplinary and may only be taken in the subject-focused pathway. Integration can grow during the research, but the question must leave room for the two subjects to meet.
2027 年新标准把总分从 34 分改成 30 分
IBO 在 2025 年 2 月出版的指南用于 2027 年首次评估，评分从旧版的 34 分改成 30 分。新版的五条标准是 A 框架 6 分、B 知识与理解 6 分、C 分析与论证线索 6 分、D 讨论与评价 8 分、E 反思 4 分。旧版的 C 批判性思考 12 分被拆成新的 C 与 D，旧的 D 呈现 4 分不再单独计分，并入 A 的结构惯例。这套变化对研究问题的影响很直接：它现在与研究方法、结构一起被 A 标准打分，权重没有提高，但连带的项目变多了。D 标准 8 分成为最重的一条，而它评的讨论与评价必须回到问题上去，问题含糊时这一条也拿不满。备考旧版资料的学生要留意两套标准的差别，别把旧的分值背进新考场。
For first assessment in 2027 the rubric totals 30 marks rather than the previous 34, with Criterion A at 6, B at 6, C at 6, D at 8 and E at 4. The old 12-mark critical thinking criterion has been split into analysis and line of argument on one side and discussion and evaluation on the other, while presentation no longer scores on its own and is folded into the framework criterion. Criterion D is now the heaviest single block, and it can only be answered against a question that is precise enough to discuss.
研究问题在引言里的写法与位置
引言的任务是把研究问题清楚地放到读者面前，位置通常在开头一两段之内，不要埋在背景叙述的第三页。写法上要把问句原样写出来，只写一次，前后用同一套措辞，不要正文里换三种说法指同一件事。问句之后要交代范围，说明研究对象、时间跨度或文本范围，让读者知道这个答案覆盖到哪里为止。接着用一两句说明为什么这个问题值得回答，以及本文用什么路径去回答，读者就能预判后面的结构。很多学生以为引言要写得像散文，结果背景写了八百词，研究问题还没出现，评分表里 A 标准的第一条直接落到中间档。引言不需要悬念，需要的是让评委在第二段就知道你要证明什么。
The introduction exists to put the question in front of the reader within the first page or two, quoted once and then referred to in exactly the same words throughout. Scope belongs right after the question, naming the object, the period or the texts the answer will cover, followed by a sentence on why the question is worth answering and how the essay will go about it. Building eight hundred words of background before the question appears costs marks under Criterion A, because a framework the reader cannot see is not a framework.
常见的研究问题失分类型
第一种常见失分是问题太宽，宽到四千词只能写概述，评分表里分析那一栏起不来。第二种是问题已经被现成答案覆盖，学生等于在复述结论，讨论与评价没有空间。第三种是两个问题用 and 连在一起，正文一半答一个，两边都不完整。第四种是价值判断型问句，比如某个做法是否公平，却没有可检验的证据标准。第五种是问题在写作过程中被悄悄改掉，引言还写着旧问句，结论却在答新的，评委一眼能看出来。错得最多的是第三种与第五种，因为它们在草稿阶段看起来很自然。
Four failure patterns account for most of the lost marks: a question so broad that four thousand words can only summarise, a question already settled by existing sources so that discussion has nowhere to go, two questions joined by and with neither half answered, and a normative question about fairness with no evidential standard behind it. The quieter fifth pattern is a question rewritten during drafting while the introduction and conclusion still argue the old one, which examiners notice immediately.
用 RRS 与三次反思会把问题磨出来
指南要求学生在研究过程中保留研究者反思空间，简称 RRS，用来记录想法、进度、时间线与论证的变化，这份材料本身不打分。三次反思会是强制的，第三次是论文完成后的简短面谈，也就是 viva voce，导师据此填写反思与进度表，简称 RPF。RPF 上是一条不超过 500 词的反思陈述，E 标准 4 分只看这一条。把研究问题每次改动的原因写进 RRS，反思陈述就有具体例子可写，不必等到最后编一段感想。很多学生以为反思只是走过场，其实 E 标准的评分描述明确要求评价性反思并给出具体例子，描述过程只能拿低档。问题改过三次的草稿往往比一次定稿的更好，前提是每次改动都留下了理由。
Students keep a researcher's reflection space, known as the RRS, to record ideas, progress and changes in the argument, and that record is not assessed directly. Three reflection sessions are mandatory, the last being the viva voce after the essay is finished, and the supervisor completes the reflection and progress form on the basis of them. The single reflective statement on that form runs to no more than 500 words and is the only evidence used for the 4 marks under Criterion E, so specific examples of what changed and why are worth far more than a summary of the timeline.
4000 词的字数规则与提交底线
IBO 的指南写明正文不超过 4000 词，另加一条不超过 500 词的反思陈述。字数统计通常包括引言、正文、结论与引文，不包括标题页、目录、图表、公式与计算、引用与参考文献、参考书目以及反思与进度表。超过上限的论文会在所有评估标准上打折扣，而不是只扣一处，所以留一两百词的余量比较稳。整篇论文对应大约 40 小时的工作量，导师的指导时间建议为三到五小时，其中包含三次反思会。文凭层面还有两条底线：扩展论文要与知识论一起最多贡献 3 分，并且学生必须拿到 D 或以上才能获得文凭。问题写得再漂亮，交不上或超字数，这些分数都不会发生。
The essay itself may not exceed 4,000 words and is accompanied by a reflective statement of no more than 500 words. The count normally covers the introduction, body, conclusion and quotations, while the title page, contents, tables, equations, references, bibliography and the reflection form stay outside it; work over the limit is compromised across all assessment criteria rather than penalised once. The whole task is reckoned at roughly 40 hours of student work with three to five hours of supervision, and a grade of D or higher is needed for the diploma, with the essay and theory of knowledge together worth up to three points.
A-Level / GCSE / IB / KS2-KS3 全科辅导，研究问题怎么收窄、论证怎么搭，可以带着草稿来谈。
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）
A-Level / GCSE / IB 全科辅导
扩展论文选题、研究问题打磨、论证结构梳理
16621398022 同微信
Visit aleveler.com
或截图后用「扫一扫 → 相册」选择本图