IB EE 研究问题 选题 聚焦 4000词 结构
Extended Essay 里最贵的四十个字，是研究问题那一句。它写在引言里，却决定了整篇的选题边界、证据类型和结论形状。新版 EE 指南从 2027 年首次评估开始启用，第一条标准就叫 Framework for the essay，考的是研究问题、研究方法和结构三件事。旧版标准用到 2026 年场次，同样把研究问题放在第一条。两道闸门看的是同一件事：这句话立得住，后面才有分可拿。下面按设计顺序拆开，从标准定位到最后的自查动作。
The most expensive sentence in an extended essay is the research question, because it sets the boundary of the topic, the type of evidence you can use and the shape of the conclusion. The 2027 criteria put it inside criterion A, Framework for the essay, together with research methods and structure. The legacy criteria used up to the 2026 sessions placed it in the same position. Either way, the sentence has to hold before the rest of the essay can score.
研究问题落在哪条标准里
新版指南把评分改成 30 分五个标准，A 是 Framework for the essay，6 分，考研究问题、研究方法与结构。旧版是 34 分五个标准，A 叫 Focus and method，同样 6 分。两版的措辞不同，看问题的角度一致：问题要相关、清楚、聚焦，还要配得上 4000 词这个篇幅。Criterion A 的 5 到 6 分档写得直白，问题要 relevant to the topic、clear、focused in relation to the scope of the essay。很多学生以为研究问题只是引言里的一句话，其实 A 标准的三个层级都从它出发。评分用 best-fit，阅卷员从最高档往下读到最贴合的一档，问题含糊，其余四条标准也很难补回来。
Under the 2027 criteria the essay is marked out of 30, and criterion A, Framework for the essay, carries 6 marks for the research question, research methods and structure. The older 2016 guide, used up to the 2026 sessions, kept the same 6 marks under the name Focus and method inside a 34 mark total. Both versions want a question that is relevant, clear and focused in relation to the scope of a 4,000 word essay. Examiners read from the top level down and stop at the best fit, so a vague question costs marks well beyond criterion A.
好问题的四条判据
IBO 的指南给出四条判据。第一层是 worthy and workable，题目要值得做，也要有够用的资料。第二层是逼着你做研究与评价，不能写成描述或叙事。第三层是 4000 词之内收得住，多线并进的问题通常写到一半就散。第四层是结论不能一眼看出，否则整篇没有论证空间。常见的错误是拿一个常识题开头，比如某场战争是否影响了某个国家，写到最后只是复述大家已知的事。把问题改成带比较或带因果的形式，才有可争的余地。指南里那句 leads to a conclusion that is not immediately obvious，值得抄在笔记本第一页。
IBO lists four tests for a workable question: it must be worthy and workable with enough sources, it must compel research, evaluation and analysis rather than description, it must fit 4,000 words without multi-pronged branches, and it must lead to a conclusion that is not immediately obvious. A question everyone can answer from general knowledge leaves nothing to argue, and the essay turns into retelling. Adding comparison or a causal link creates room for judgement. That phrase about a conclusion that is not immediately obvious is worth copying onto the first page of your notebook.
把宽泛兴趣收成一道题
从兴趣到问题通常要动三处，对象、时间与地点或变量。以经济为例，兴趣是「最低工资好不好」，问题要收成「某国 2021 到 2025 年最低工资上调对 18 至 24 岁就业率的影响」。收窄的标尺不是字数，而是能不能在 4000 词里把证据铺完。学生最常把范围收得太窄，落到一个只有一篇论文谈过的样本，资料不够，方法也做不出比较。可行的顺序是先列 8 到 10 个候选子问题，删掉资料拿不到的，剩下的按可争论程度排序。第二步才是写正式问句，中文与英文各写一遍，看两边意思有没有偏。阅卷时能看出的偏移，往往在英文表述里先露出来。
Moving from interest to question usually means narrowing three things: who or what, when, and where or which variable. An interest in whether minimum wages are good becomes a question about the effect of a specific increase on 18 to 24 year old employment in one country between 2021 and 2025. The test is not word count but whether the evidence can be laid out inside 4,000 words, and narrowing too far leaves one paper and no comparison. Write eight to ten candidate sub-questions, drop the ones without accessible sources, then turn the survivor into a full sentence in both languages and check that the two versions say the same thing.
疑问词决定论证方式
How far、To what extent、How does X affect Y、Should 这几类开头，对应的论证结构完全不同。To what extent 要求给出程度判断并说明边界，结论必须回答在多大程度上成立。How does X affect Y 要拆机制，机制讲不清就只剩相关性。Should 属于规范性判断，需要先立评价标准，再拿证据对照。很多学生以为疑问词只是语气差别，其实它决定了每一节该写什么。把 To what extent 写成是否题，结论会空；把是否题写成程度题，证据又不够。改疑问词往往比改内容省时间。
The opening phrase of a question fixes the shape of the argument. To what extent demands a judgement of degree with stated limits, so the conclusion has to say how far the claim holds. How does X affect Y demands a mechanism, and without one the essay is left with correlation. Should questions are normative, which means an evaluation standard has to be established before evidence is weighed; changing the opening phrase is usually faster than rewriting the content beneath it.
问题要能被证据回答
一道题能不能做，取决于数据是否拿得到。历史题要问档案与一手材料是否存在，科学题要问实验条件是否具备，文学题要问文本有没有足够的解读空间。常见的错误是先定问题再找资料，结果发现关键数据不公开，只好把论证改成推测。反向验证更省事，先列出三类能拿到的来源，再看问题能不能被这三类来源支撑。旧版标准在 Criterion A 里写明 source(s) and/or method(s) 要 an appropriate range。新版写成 research methods that are suitable for the research question are explained and applied effectively。两处都在说同一件事，方法与问题要对得上。
Whether a question can be answered depends on access to data, not on how interesting it sounds. Historians need archives, scientists need equipment and repeatable conditions, literature students need texts with enough interpretive space, and an economics essay needs reports and figures rather than textbooks alone. Fixing the question first and hunting sources afterwards usually ends in speculation. Work backwards instead: list three source types you can realistically obtain, then test whether they can carry the question, because both versions of criterion A ask for methods that suit the question.
学科归属与那道分数天花板
题目必须在注册科目范围内，新旧两版都给了硬后果。旧版指南写明，如果题目或研究问题被认为不适合所注册的科目，A 与 B 两个标准最多只能给 4 分，C 最多 3 分。一道写得很顺的题，只要学科归属错了，分数上限立刻压下来。常见的错误是把心理学题写成社会学评论，或者把经济学题写成商业案例分析。判断方法很直接，看问题里用到的方法与术语是不是该科目的核心工具。拿不准就问导师，IBO 的科目指南对每个科目能接受什么题目有明确说明。
A question must sit inside the subject the essay is registered in, and both guides attach a hard ceiling to getting this wrong. Under the legacy criteria, a topic or question judged inappropriate for the registered subject caps criterion A and B at 4 marks and criterion C at 3, so a fluent essay can still lose heavily. Psychology questions rewritten as sociology commentary, or economics questions turned into business case studies, are the usual examples. The practical test is whether the methods and terminology in the question are the core tools of that discipline, and IBO subject guidance sets out what each subject accepts.
研究方法要写进问题
研究问题里最好能看出一条方法线索，实验、问卷、文本分析还是二手数据。新版指南把方法放在 A 标准的第一层，要求它与问题一起被解释。写「空气质量如何影响某地学生的出勤率」，方法自然是数据比对与相关分析。写「某部小说的叙事视角如何制造不可靠叙述」，方法就是文本细读。方法不明的题，写到中段会突然发现每一节都在换工具，前后无法对照。导师会谈时可以直接问一句，这个问题的证据从哪来，答不上来就说明题还没成形。
A good question leaves a visible trail to its method. Experimental work, surveys, textual analysis and secondary data all imply different evidence, and the 2027 criterion A asks for methods that are suitable for the question and effectively applied. A question about air quality and attendance points to data comparison and correlation, while a question about narrative perspective points to close reading. When the method is missing, the middle of the essay starts swapping tools section by section, so a single question in the first supervision meeting settles it: where does the evidence come from?
三个坏问题的拆解
太宽的典型是「社交媒体是否影响青少年心理健康」，一句话能写一本书。可辩论但没有证据的典型是「某政策是否公正」，标准立不住就只能空谈。描述性的典型是「某战役的过程与结果」，写出来是复述。三种坏题的共同点是没有可检验的判断。改法分别是加限定与变量、先立评价标准、把过程题改成因果或影响题。改完回到四条判据上过一遍，值得做、逼着分析、4000 词内、结论不显然。
Three failures cover most weak questions. Too broad looks like asking whether social media affects adolescent mental health, which is a book rather than an essay; arguable but unevidenced looks like asking whether a policy was just, with no standard to measure justice against. Descriptive looks like retelling the course of a battle. None of them carries a testable judgement, and the three fixes are adding limits and variables, establishing an evaluation standard first, and turning process into cause and effect.
用问题检查全文聚焦
引言要把问题写清楚，正文每一节都要回应它，结论要直接回答它。旧版 A 标准的措辞是 how the focus is maintained throughout the essay。新版写成 structural conventions used in the essay effectively support communication of the research。两句话都在要求一条不断的线。实操办法很朴素，每写完一节，在旁边写一句这一节回答了问题的哪一部分，写不出来的节通常是多余。删掉冗余比补写快。4000 词的上限之下，多出来的每一段都在挤掉别的证据。
The introduction states the question, every section answers part of it, and the conclusion answers it directly. The legacy descriptors call this maintaining focus throughout the essay, and the 2027 version asks whether structural conventions support communication of the research. One habit keeps the line unbroken: after drafting a section, write in the margin which part of the question it answers, and delete sections with no answer. Under a 4,000 word limit, every redundant paragraph pushes out evidence that could have earned marks.
研究问题与三次反思
反思表不是走过场，E 标准只按 RPPF 的前 500 词评。旧版 E 叫 Engagement，6 分。新版改成 Reflection: evaluative; growth，4 分，看的是你有没有评价自己的学习过程与迁移。第一次反思对应选题与初步问题，第二次对应研究过程中的调整，第三次对应结论与思考变化。问题改过几次、为什么改，是这份表里最有说服力的内容。很多学生以为反思要写感受，其实写决策更得分。把一次实质性收窄写清楚，比写十句努力更有效。
The reflections on planning and progress form are assessed, not decorative, and only the first 500 words count under criterion E. The legacy version gave engagement 6 marks, while the 2027 version gives reflection on evaluative practice and growth 4 marks and looks for specific examples and transfer of learning. The first reflection belongs to the topic and initial question, the second to adjustments during research, the third to the conclusion and what shifted in your thinking. Accounts of how the question narrowed are the strongest material, because they show decisions rather than effort.
第一次导师会谈该问的四件事
会谈不是汇报进度，是把问题磨到能用。可以带着四个问题去，这个问题在本科目里成立吗，关键证据在哪，4000 词够不够铺完，结论有没有可能被反驳。导师通常不会替你改题，但会指出哪一处范围过大。会后把讨论整理成第一次反思的初稿，控制在这张表 500 词上限之内。旧版指南把反思会议定为三次强制会谈，第一次在开始研究之前，第二次在中期，第三次在交稿前。每一次都留书面记录，后面写 RPPF 就不用靠回忆。
A supervision meeting is for sharpening the question, not reporting progress. Bring four things: does this question belong in my subject, where is the key evidence, can it be covered in 4,000 words, and could the conclusion be refuted? Supervisors rarely rewrite a question for you, but they will point at the part that is too wide. Turn the discussion into a draft of the first reflection while it is fresh, keeping the three reflections inside the 500 word limit on the form; the legacy guide made three reflection sessions mandatory, before research, at the midpoint and before submission.
交稿前要过的四处扣分点
标题页不写标题是常见疏漏，新版指南把标题列为正式要求，缺失会与结构一起在 D 标准里算账。问题比标题还笼统也很常见。第三种是整篇不提问题，只在引言出现一次。第四种是问题改过但正文没跟上，前后两套说法并存。阅卷员按标准逐条找证据，找不到就按低档给分。交稿前做一遍机械检查，把引言里那句话抄出来，翻到每一节看有没有对应，再看结论有没有正面回答。这一步十分钟，能省下的分数不少。
Missing the title on the title page is a common omission, and the 2027 guide treats it as a formal requirement weighed with structure under criterion D. Other frequent losses are a question vaguer than the title, a question that appears once in the introduction and never again, and a revised question that never reached the body so two versions coexist. Examiners look for evidence criterion by criterion and mark low where they find none. A ten minute check before submission settles it: copy the sentence from the introduction, find its counterpart in each section, then confirm the conclusion answers it.
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
需要一对一辅导？Need One-on-One Tutoring?
A-Level / GCSE / IB / KS2-KS3 全科辅导，EE 选题与写作全程跟进
16621398022 同微信
Visit aleveler.com
或截图后用「扫一扫 → 相册」选择本图