📚 Postgraduate Coursework and Research Tutoring | 研究生课程及科研辅导
Postgraduate study changes the rules of the game. On a master's or doctoral programme you are no longer rewarded mainly for reproducing a syllabus; you are judged on how well you handle primary sources, frame a researchable question, justify a method, analyse evidence and defend your conclusions in writing and in person. Many capable students find that it is this transition - from consuming knowledge to producing it - where the real difficulty begins. Structured one-to-one tutoring exists to make that transition manageable.
研究生阶段的学习规则与本科完全不同。在硕士或博士项目中，成绩不再主要取决于你能多么准确地复述教学大纲，而是取决于你能否处理一手文献、提出可研究的问题、为研究方法提供依据、分析证据，并以书面和口头方式为自己的结论辩护。很多能力很强的学生发现，真正吃力之处正是这种从"接受知识"到"生产知识"的转变。有结构的一对一辅导，就是让这个转变变得可控。
1. Who Needs Postgraduate Coursework and Research Tutoring | 谁需要研究生阶段的课程与科研辅导
Students who come to us usually fall into a handful of recognisable groups: taught master's students struggling with unfamiliar quantitative, statistical or coding content; research master's students preparing a proposal, an upgrade paper or a first-year report; dissertation and thesis writers who need structure, feedback and deadlines; and doctoral candidates who need help turning a broad interest into a focused, publishable study.
来找我们辅导的学生大致分为几类：修读授课型硕士、对定量方法、统计或编程内容不熟悉的学生；正在准备研究计划书、转博论文或一年级报告的研究型硕士；需要框架、反馈与时间节点的学位论文写作者；以及需要把宽泛兴趣转化为聚焦且可发表研究的博士生。
A second group are students studying in a second language. Writing a literature review, chairing a supervision meeting or answering questions in a viva demands a level of academic English that general language courses rarely reach. Here the tutoring is not about grammar drills; it is about register, hedging, signposting and the specific rhetorical moves that examiners in the UK, Australia, Canada and the US expect to see.
第二类是在非母语环境中学习的学生。撰写文献综述、主持一场导师例会、在答辩中回答问题，所需要的学术英语水平，普通语言课程往往达不到。这类辅导不是做语法练习，而是聚焦语域、缓和表达、行文衔接，以及英、澳、加、美考官明确期待的修辞动作。
Finally, there are students who are not in difficulty at all. They are doing well but want to move faster - to publish earlier, to win a competitive scholarship, or to convert a dissertation chapter into a conference paper. Tutoring works just as well as acceleration as it does as rescue.
最后还有一类学生其实并没有遇到困难。他们表现良好，只是想走得更快——更早发表、拿到竞争激烈的奖学金，或者把学位论文的一章转化为会议论文。辅导既能"救急"，也能"加速"。
2. Taught Master's, Research Master's and PhD: Different Needs | 授课型硕士、研究型硕士与博士：需求各不相同
A taught master's (MSc, MA, MEd, MBA) is assessed mainly through modules, examinations and a final project of 10,000-20,000 words. Support here is module-led and calendar-driven: coursework deadlines, problem sets, lab reports, presentations and the summer dissertation.
授课型硕士（MSc、MA、MEd、MBA）主要通过模块考试和一个 10,000-20,000 字的毕业项目评估。这类辅导以模块为单位、以日程为驱动：平时作业截止日期、习题集、实验报告、展示汇报，以及暑期的学位论文。
A research master's (MRes, MPhil) sits between coursework and a doctorate. The core deliverable is usually a proposal followed by a substantial thesis, and the hardest part is often the first six months: narrowing a topic, mapping the field and producing a defensible research question.
研究型硕士（MRes、MPhil）介于授课与博士之间。核心产出通常是研究计划书加一篇有分量的学位论文，而最艰难的部分往往是前六个月：收窄题目、梳理领域地图、形成一个站得住脚的研究问题。
Doctoral work adds two demands that master's study does not: originality and continuity. A PhD must make an independent contribution to knowledge, and it must keep going for three to four years without external structure. That is why so much doctoral tutoring focuses on project management and on the "upgrade" or "confirmation" milestone as much as on the research itself.
博士阶段增加了硕士阶段没有的两项要求：原创性与持续性。博士必须对知识作出独立贡献，并且要在没有外部结构的情况下持续三到四年。这也是为什么博士阶段的辅导，很大一部分精力既在项目管理和"转博/资格确认"节点上，也在研究本身。
3. Module-Level Support for Taught Postgraduate Courses | 授课型研究生课程的模块辅导
Most taught postgraduate programmes fail students not because the ideas are too hard but because the assessment format is unfamiliar. A student who has never written an econometrics report, a systematic review protocol or a reflective portfolio can lose marks for structural reasons rather than intellectual ones.
大多数授课型研究生项目让学生失分，并不是因为概念太难，而是因为评估形式陌生。一个从未写过计量经济学报告、系统综述方案或反思性作品集的学生，可能因为结构问题而非智力问题丢分。
Module-level tutoring therefore begins with the assessment itself: the marking rubric, the learning outcomes and the previous year's paper. From there we work backwards to the content that actually needs to be understood, rather than re-teaching an entire textbook.
因此，模块辅导从评估本身入手：评分细则、学习成果，以及上一年的考题。然后我们倒推回去，只处理真正需要理解的内容，而不是把整本教材重讲一遍。
| Module type | Typical difficulty | Focus of tutoring |
| Statistics / econometrics | Interpreting output, choosing tests | Assumption checks, model fit, write-up |
| Research methods | Linking paradigm to method | Design logic, sampling, validity |
| Coding / data science | Debugging, reproducibility | Python, R, SQL, clean pipelines |
| Theory-heavy modules | Reading density, seminar discussion | Argument mapping, critical synthesis |
| Dissertation / project | Scope, structure, momentum | Chapter planning, staged feedback |
统计类模块：难点在解读软件输出、选择合适的检验；辅导聚焦前提假设检验、模型拟合与结果写作。研究方法模块：难点在把研究范式与方法对应起来；辅导聚焦设计逻辑、抽样与效度。编程与数据科学模块：难点在调试与可复现性；辅导覆盖 Python、R、SQL 与干净的数据流程。理论密集模块：难点在阅读量与研讨发言；辅导聚焦论证结构图与批判性综合。学位论文模块：难点在范围、结构与节奏；辅导聚焦章节规划与分阶段反馈。
4. Literature Search, Reading and Critical Appraisal | 文献检索、阅读与批判性评估
The first serious skill of postgraduate work is finding the right literature and reading it efficiently. That means knowing your databases (Scopus, Web of Science, PubMed, JSTOR, PsycINFO, IEEE Xplore), building reproducible search strings with Boolean operators, and keeping a reference library that does not collapse halfway through the thesis.
研究生阶段第一项硬技能，是找到对的文献并高效阅读。这包括熟悉所在领域的数据库（Scopus、Web of Science、PubMed、JSTOR、PsycINFO、IEEE Xplore），能用布尔逻辑构建可复现的检索式，并维护一个不会在论文进行到一半就崩掉的文献库。
Efficient reading is selective. A useful routine is to read the abstract, then the conclusion, then the figures and tables, and only then decide whether the whole paper deserves your time. When you do read in full, annotate for four things only: the claim, the evidence, the method, and the limitation.
高效阅读的关键是选择性。一个实用的流程是：先读摘要，再读结论，然后看图表，最后才决定这篇论文是否值得投入完整时间。当你确实要通读时，只做四类批注：主张、证据、方法和局限。
Critical appraisal means asking whether the study could have found anything else. Was the sample adequate? Was the measure valid? Was the analysis pre-specified or exploratory? Were the authors' conclusions proportionate to their data? Examiners reward students who can answer these questions about someone else's work before applying the same scrutiny to their own.
批判性评估的意思是问：这项研究有没有可能得出别的结论？样本量是否充分？测量工具是否有效？分析是事先设定的还是探索性的？作者的结论与数据是否匹配？考官格外欣赏那些能先对他人的研究提出这些问题、再把同样标准用在自己身上的学生。
5. Writing a Strong Literature Review | 写出有说服力的文献综述
A literature review is not a list of summaries. It is an argument about the state of a field: what is established, what is contested, what is missing, and why your study is the logical next step. The most common failure is the "annotated bibliography" structure, in which each paragraph describes one paper and nothing connects them.
文献综述不是摘要清单，而是关于一个领域现状的论证：什么已经确立、什么仍有争议、什么尚属空白，以及为什么你的研究是逻辑上的下一步。最常见的失败是把它写成"注释书目"——每段介绍一篇文献，彼此之间没有连接。
A workable structure moves from the general to the specific across three or four thematic sections, each ending with a synthesis sentence that states what the body of work collectively shows. Tenses are used deliberately: the past tense for what a single study did, the present tense for what the field now holds to be true.
一个可行的结构是在三到四个主题板块中由宽泛走向具体，每个板块以一句综合句结尾，说明这一批研究整体上说明了什么。时态要有意识地使用：单篇研究做了什么用过去时，领域当前公认的结论用现在时。
Then comes the gap statement. It must be precise and defensible - not "little research exists on X" (which is usually false) but "existing studies of X have relied on cross-sectional designs and self-report measures, so causal direction remains unclear". The gap you name must be the gap your study closes.
接下去是"研究空白"陈述。它必须精确且站得住脚——不是"关于 X 的研究很少"（这通常是错的），而是"现有关于 X 的研究多采用横断面设计与自陈量表，因此因果方向仍不明确"。你指出的空白，必须正是你的研究要填补的空白。
6. Research Design and Methodology | 研究设计与方法论
Methodology is where most postgraduate projects are won or lost. The design must follow from the question, not from the methods the student happens to know. If the question is about prevalence, you need a survey; if it is about mechanism, you probably need an experiment or a longitudinal design; if it is about meaning, you need qualitative inquiry.
方法论是大多数研究生项目成败的关键。设计必须由研究问题推导出来，而不是由学生恰好掌握的方法决定。如果问题关乎流行程度，你需要问卷；如果关乎机制，你多半需要实验或追踪设计；如果关乎意义，你需要质性探究。
At the heart of the design sit a few concepts that examiners always test: internal validity (does the design support a causal claim?), external validity (to whom does the result generalise?), reliability (would the measure give the same answer again?) and construct validity (are you measuring the thing you named?).
设计的核心是几个考官必然会考的概念：内部效度（该设计是否支持因果主张？）、外部效度（结果可以推广到谁？）、信度（同一测量再测一次是否得到同样答案？）、构念效度（你测量的究竟是不是你所命名的那个东西？）。
Ethics approval, sampling strategy, sample size justification and piloting belong in this chapter, not in an appendix. A power calculation, for example, is not decoration: it is the argument that your study is capable of detecting the effect you claim to be looking for.
伦理审批、抽样策略、样本量论证与预调查，都属于这一章，而不是放进附录。例如统计功效计算不是装饰品，而是"本研究有能力侦测到你所声称要寻找的效应"这一论点的支撑。
7. Quantitative Analysis and Statistical Support | 定量分析与统计辅导
Statistical tutoring usually starts with the data, not the software. Before any test is run, we check the structure: how many cases and variables, what is missing, what the distributions look like, and whether the measurement level of each variable matches the test being planned.
统计辅导通常从数据本身开始，而不是从软件开始。在跑任何检验之前，先检查数据结构：有多少个案与变量、缺失情况如何、分布形态如何，以及每个变量的测量层次是否与计划使用的检验相匹配。
interpret the effect size rather than the p-value alone. A result reported as t(48) = 2.31, p = 0.025, d = 0.66 says far more than "the difference was significant".
从这里出发的逻辑很清晰：根据研究问题和数据类型选择检验方法，检查该方法所需的前提假设，运行检验，然后解读效应量而不只看 p 值。写作 t(48) = 2.31，p = 0.025，d = 0.66 这样的结果，信息量远大于一句"差异显著"。
Common topics include t-tests, ANOVA and its non-parametric alternatives, correlation and regression, logistic regression, factor analysis, multilevel models, and the assumptions that sit under each: normality, homogeneity of variance, independence, multicollinearity and linearity. Output from SPSS, Stata, R or Python is translated into sentences that belong in a results chapter.
常见主题包括 t 检验、方差分析及其非参数替代方法、相关与回归、逻辑回归、因子分析、多层模型，以及各自背后的假设：正态性、方差齐性、独立性、多重共线性与线性关系。SPSS、Stata、R 或 Python 的输出，会被翻译成能直接写进结果章节的句子。
We also spend time on what not to do: p-hacking, dropping outliers without justification, reporting only the models that "worked", and describing correlation as causation. These are the mistakes that damage a thesis and, occasionally, a career.
我们同样会花时间讲清楚哪些事不能做：p 值操纵、无理由剔除异常值、只报告"跑通了"的模型，以及把相关当作因果。这些错误会损害一篇学位论文，有时甚至损害一段学术生涯。
8. Qualitative Analysis and Coding | 定性分析与编码
Qualitative work carries its own rigour. Whether you use thematic analysis, grounded theory, IPA, discourse analysis or case study design, the examiner wants to see a transparent audit trail from raw material to interpretation.
质性研究有其自身的严谨性标准。无论你采用主题分析、扎根理论、解释现象学分析（IPA）、话语分析还是案例研究设计，考官都希望看到一条从原始材料通往解释的透明可追溯链条。
A practical workflow looks like this: transcribe verbatim, read the whole corpus once without coding, generate initial codes close to the data, cluster codes into candidate themes, review those themes against the full dataset, then define and name them. Each stage is documented, and any deviation from the stated method is acknowledged.
一个可操作的工作流程是：逐字转写；先不带编码地通读全部语料；生成贴近数据的初始编码；把编码聚类为候选主题；回到全部数据检验主题；最后定义并命名主题。每个阶段都要留档，任何对既定方法的偏离都要说明。
Reflexivity is not a formality. A short, honest statement about your position relative to the participants and the topic strengthens the analysis. Quotations should be chosen because they illustrate a theme, not because they sound eloquent, and every theme should be supported by more than one participant where the sample allows.
反身性不是走过场。简短而诚实地说明你相对于被访者与议题的位置，会增强分析的可信度。引文的选择标准是它能否说明某个主题，而不是它读起来是否漂亮；在样本允许的情况下，每个主题都应由不止一位被访者的材料支撑。
9. Academic Writing, Referencing and Integrity | 学术写作、引用与学术诚信
Academic writing at postgraduate level is judged on clarity, precision and appropriate caution. Long sentences are not a sign of sophistication. Good academic prose usually states the point first, then supports it, then qualifies it.
研究生层次的学术写作，评判标准是清晰、精确与适度的审慎。句子长并不代表水平高。好的学术写作通常是先陈述观点，再给出支撑，最后作限定说明。
Referencing must be consistent and complete, in whichever style your programme requires - APA 7th, Harvard, MLA, Chicago, OSCOLA or Vancouver. Reference management software (Zotero, EndNote, Mendeley) saves time, but it does not remove the obligation to check that every in-text citation appears in the list and every listed source is cited.
引用必须一致且完整，具体采用所在项目要求格式——APA 第 7 版、Harvard、MLA、Chicago、OSCOLA 或 Vancouver。文献管理软件（Zotero、EndNote、Mendeley）能节省时间，但并不能免除你的责任：必须核对每一条文中引用都出现在文末列表中，且列表中每一条都被引用过。
On integrity, the line is clear. Quoting without quotation marks, paraphrasing without citation, reusing your own previously submitted work without declaration, and submitting text produced by a generative tool as your own analysis are all breaches. Generative tools may be used for brainstorming or language checking only where your institution permits it, and their use should be declared.
在学术诚信上，界线是清楚的。引用不加引号、改写不加出处、未经声明重复使用自己此前提交过的作业、把生成式工具产出的文本当作自己的分析提交，都属于违规。只有在所在院校允许的前提下，生成式工具才可用于头脑风暴或语言检查，且使用情况应当声明。
10. Working Effectively With Your Supervisor | 高效与导师合作
Supervision is a professional relationship with its own etiquette. Supervisors are busy, they supervise several students, and they respond to specificity. "I am stuck" wastes a meeting; "I have drafted three possible research questions and I think option 2 is strongest because the data are accessible - do you agree?" uses fifteen minutes well.
导学关系是一种有自身礼仪的职业关系。导师很忙，同时带若干学生，而且他们只对具体的问题给出具体回应。"我卡住了"会浪费一次例会；"我起草了三个可能的研究问题，我认为第二个最强，因为数据可得——您同意吗？"才是对十五分钟的有效利用。
Practical habits help: send an agenda and any draft at least 48 hours before the meeting; keep a running log of decisions made; agree on the next deliverable and its date before you leave the room; and follow up with a short summary email so that expectations are on the record.
一些实用习惯会有帮助：至少在会面前 48 小时发出议程与稿件；维护一份持续更新的"已定事项"记录；离开会议室之前确定下一项交付物及其日期；会后发一封简短总结邮件，让双方预期有据可查。
Feedback should be received as data, not as a verdict. When a supervisor writes "unclear" in the margin, the useful question is not "do they dislike my writing?" but "what exactly was unclear to a reader who does not share my assumptions?" That reframing alone can save months.
反馈应当被当作信息，而不是判决。当导师在页边写下"不清楚"时，有用的问题不是"他们是不是不喜欢我的写作"，而是"对于一个不共享我预设的读者来说，究竟哪里不清楚"。仅仅这一种重新表述，就足以省下几个月的时间。
11. Dissertation, Thesis and Viva Preparation | 学位论文与答辩准备
The dissertation is a project before it is a document. That means a plan with milestones, a word budget per chapter, and a realistic writing routine. A common and workable budget for a 20,000-word master's dissertation is roughly 3,000 words for the introduction, 6,000 for the literature review, 4,000 for methodology, 5,000 for results and 2,000 for discussion, with the remainder for the conclusion and references.
学位论文首先是一个项目，其次才是一份文档。这意味着要有带里程碑的计划、每章的篇幅预算，以及可执行的写作节奏。一篇 20,000 字硕士论文的常见可行预算大致是：引言 3,000 字、文献综述 6,000 字、方法论 4,000 字、结果 5,000 字、讨论 2,000 字，其余留给结论与参考文献。
Structure carries marks. Each chapter should open with a short roadmap and close with a link forward. The results chapter should report without interpreting; the discussion chapter should interpret, compare with the literature, acknowledge limitations and state implications. Mixing the two is one of the most frequent structural penalties.
结构本身就会带来分数。每一章都应以简短的路线图开篇，以承接下一章的句子收尾。结果章节只报告、不解释；讨论章节负责解释、与文献对话、坦承局限并说明意义。把这两者混在一起，是最常见的结构性扣分点之一。
For the viva, preparation means three things: being able to summarise the whole thesis in three minutes, at one minute and in one sentence; knowing the weaknesses of your own work better than the examiners do; and having a calm, honest answer ready for the question "what would you do differently?" Rehearsing aloud with someone who will interrupt you is far more useful than reading your chapters again.
至于答辩，准备意味着三件事：能在三分钟、一分钟和一句话内概括整篇论文；比考官更清楚自己研究的弱点；对"你会怎么做不同"这个问题准备一个冷静而诚实的回答。找一位会打断你的人大声演练，远比再读一遍自己的章节有用。
12. Study Planning, Wellbeing and How to Contact Us | 学习规划、身心状态与联系方式
Postgraduate study is a marathon with deadlines. Sustainable progress comes from a weekly rhythm rather than heroic all-nighters: fixed writing blocks, a realistic daily word or task target, one day off, and a written record of what was actually done. Progress you can see is the best defence against the isolation that research often brings.
研究生学习是一场带截止日期的马拉松。可持续的进展来自每周的节奏，而不是偶尔的通宵突击：固定的写作时段、切实可行的每日字数或任务目标、每周留出一天休息，以及一份"实际完成了什么"的书面记录。看得见的进展，是抵御研究常带来的孤独感的最好方式。
If you are struggling, say so early. Extensions, mitigating circumstances procedures and student wellbeing services exist precisely for this, and using them is not a failure. Burnout costs far more time than a two-week extension ever will.
如果你正陷入困境，请尽早说出来。延期申请、特殊情况（mitigating circumstances）程序与学生心理健康服务的存在，正是为了应对这种情况，使用它们并不等于失败。倦怠
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