威威老師的 GRE Analytical Writing 完整攻略 ✍️🌍
課程導航
哈囉各位同學！我是威威老師。
來，先給你警示：
「我 Verbal 考 158、Quant 考 165——AWA 隨便寫就好？」
🚨 錯！大錯特錯！
美國研究所的招生委員會非常看重 AWA 分數——因為它直接反映你的「學術論證能力」！
- Verbal/Quant 高分 → 證明你會考試
- AWA 高分 → 證明你能寫論文
就算 Verbal 158、Quant 165——AWA 只有 2.5 分→教授會質疑你真的能寫論文嗎？
💡 威威鐵律： AWA 4.0+ 是基本，5.0+ 才有競爭力！
今天威威老師帶你拆解：
- 🎯 AWA 目前格式（2023 改版後）——只考 Issue Essay 一篇
- 📋 5 段論證結構模板——背了就用
- 🔥 批判思考要素——complexity、specific examples、logical flow
- ⚡ 30 分鐘時間管理——5 分讀題 + 20 分寫 + 5 分檢查
繫好安全帶 🚂
一、考試格式總覽
🚨 2023 年 9 月改版後： AWA 只剩 1 題 Issue Essay（30 分鐘）！
| 項目 | 細節 |
|---|---|
| Essay 數量 | 1 篇（改版後不再考 Argument Essay） |
| 時間 | 30 分鐘 |
| 題型 | Analyze an Issue |
| 計分方式 | 0.0-6.0 分，0.5 分為單位 |
| 評分者 | 1 位 human rater + 1 台 e-rater（電腦評分） |
| 寫作方式 | 打字（鍵盤輸入）——無法手寫！ |
評分標準（Holistic Scoring）
| 分數 | 等級 | 描述 |
|---|---|---|
| 6.0 | Outstanding | 深度洞察力、論證極具說服力、語言精準優美 |
| 5.0 | Strong | 思路清晰有深度、論證充分、語言流暢 |
| 4.0 | Adequate | 有基本的論證和分析、語言尚可、有些許問題 |
| 3.0 | Limited | 論證或分析有明顯缺陷、語言能力有限 |
| 2.0 | Seriously Flawed | 分析嚴重不足、語言問題多 |
| 1.0 | Fundamentally Deficient | 基本上沒有內容 |
| 0.0 | No Score | 離題、非英文作答、沒打字 |
常見錯誤
很多台灣學生把 Issue Essay 當成「英文作文比賽」——拼命用高級單字和複雜句型。但 GRE 寫作考的是分析邏輯，不是文學創作！一篇用詞簡單但邏輯嚴謹的文章，分數會遠高於詞藻華麗但論證空洞的文章。請把精力放在「想清楚」而不是「寫漂亮」！
二、Issue Essay（議題寫作）
2.1 什麼是 Issue Essay？
Issue Essay 要求你就一個具有爭議性的議題表達立場並進行論證。題目形式通常是一句簡短的陳述（claim），你必須：
- 明確表達你同意或不同意（或部分同意）的程度
- 用具體的理由和例子支持你的立場
- 考慮反方觀點並回應
考試會給你一個題庫中的題目（隨機抽取），你可以從題庫中準備！ ETS 官網有完整 Issue Pool（約 150+ 題），考前務必瀏覽一遍。
2.2 評分標準（Issue Essay 四維度）
| 維度 | 評估什麼 | 台灣學生常見問題 |
|---|---|---|
| Position（立場） | 是否有清晰、明確的立場？ | 立場模糊、兩邊都不得罪 → 主論點不明 |
| Development（發展） | 用了哪些理由和例子？是否充分？ | 有主張但沒有具體例子 → 空洞 |
| Organization（組織） | 結構是否清晰、邏輯是否連貫？ | 想到什麼寫什麼 → 跳躍式思維 |
| Language（語言） | 句型多樣性、用詞精準度、文法正確性 | 過度使用簡單句／中式英文 |
2.3 Essay 結構模板與句型框架
標準五段式結構
Paragraph 1: Introduction（引言）
- Hook + Context + Thesis Statement
Paragraph 2: Body 1（主論點一 + 例子）
- Topic Sentence + Example + Explanation
Paragraph 3: Body 2（主論點二 + 例子）
- Topic Sentence + Example + Explanation
Paragraph 4: Body 3（反駁反方觀點／讓步後反駁）
- Concession + Rebuttal
Paragraph 5: Conclusion（結論）
- Restate Thesis + Summary + Final Thought
句型框架（Sentence Frames）
Introduction（引言段）
The statement that [簡述題目主張] raises a complex question about [核心議題].
While there is some merit to this perspective, I largely [agree/disagree] with the claim for [數字] key reasons.
OR:
The assertion that [簡述題目主張] is one that has engendered considerable debate.
I find this position [persuasive/unpersuasive] to a significant extent, as will be demonstrated through an analysis of [領域一] and [領域二].Body Paragraphs（主體段）
First and foremost, [topic sentence — 本段的具體主張].
Consider, for example, [具體例子].
This case illustrates that [解釋例子如何支持你的主張].
If [反方可能說的話], one might counter that [回應反方].
Furthermore, [第二個具體主張].
A compelling illustration can be found in [第二個例子 — 最好來自不同領域].
The significance of this example lies in [說明這個例子證明了什麼].Concession + Rebuttal（讓步＋反駁）
Admittedly, proponents of [反方立場] might argue that [反方最強論點].
This concern is not without foundation; indeed, [承認反方有某些道理].
However, this objection ultimately fails to undermine the broader argument because [解釋為什麼你的立場仍然成立].
The [反方的擔憂] can be addressed through [解決方案] rather than [極端方案].Conclusion（結論段）
In conclusion, while the claim that [原題目] contains a grain of truth,
it [overstates the case/overlooks crucial countervailing factors].
The examples from [領域一] and [領域二] demonstrate that [你的核心觀點].
Ultimately, [最後一句總結——可以是建議、預測、或更廣泛的反思].2.4 如何快速產出具體例子
這是台灣學生最大的痛點：想不到例子可以寫！以下是威威老師的「四象限例子生成法」：
| 象限 | 領域 | 例子類型 | 範例 |
|---|---|---|---|
| Personal | 個人經驗 | 你身邊的故事 | 台灣的教育經驗、職場觀察 |
| Historical | 歷史事件 | 眾所周知的歷史事實 | 工業革命、二戰、台灣民主化 |
| Literary/Cultural | 文學／文化 | 經典作品或文化現象 | Shakespeare、村上春樹、社群媒體興起 |
| Scientific/Technological | 科學／科技 | 科學發現或科技趨勢 | AI 發展、氣候變遷研究、疫苗開發 |
威威老師小提醒
GRE 評分者不在乎你的例子「夠不夠學術」——他們在乎例子是否具體且能支持論點。與其寫一個模糊的歷史事件，不如寫一個你親身經歷的小故事，只要能清楚說明道理就好！考前可以準備 5-8 個「萬用例子」（universal examples），例如：AI 發展、氣候變遷、COVID-19 疫情、社群媒體的影響——這些幾乎可以套用在任何議題上。
2.5 時間分配策略
| 階段 | 時間 | 做什麼 |
|---|---|---|
| Plan（計畫） | 5 分鐘 | 分析題目、決定立場、列兩個理由＋一個反駁、各配一個例子 |
| Write（寫作） | 22 分鐘 | 五段式文章，每段約 4–5 分鐘 |
| Review（檢查） | 3 分鐘 | 檢查拼字、文法、確認每段有 topic sentence |
2.6 五大常見 Issue 題目 + 分析 + 範文
Issue Topic 1: Education（教育）
Prompt: “To be an effective leader, a public official must maintain the highest ethical and moral standards.”
分析題目：
- 核心議題：領導者的道德標準是否必須是「最高」的？
- 關鍵詞：“effective”（有效果的——不是受歡迎的）、“highest”（最高標準——是絕對標準嗎？）
- 可採立場：部分同意——道德很重要但不是唯一要素；效能（competence）同樣重要
Brainstorm 雙面觀點：
| 同意方（Agree） | 反對方（Disagree） |
|---|---|
| 道德敗壞會摧毀公眾信任 | 有些極有效的領導者道德有瑕疵（例如某些革命領袖） |
| 公眾人物的行為有示範效應 | 過於道德潔癖可能導致決策猶豫不決 |
| 長期來看誠信是不可或缺的 | 「有效」不等於「道德」——實用主義有時更有效 |
Model Essay（約 450 words）：
The assertion that effective leadership in public office requires the “highest” ethical and moral standards presents a deceptively simple ideal. While I agree that integrity is a crucial component of legitimate governance, I contend that this statement, taken literally, sets an unrealistic and potentially counterproductive benchmark. The relationship between moral purity and political effectiveness is, in reality, far more nuanced.
First and foremost, the demand for the “highest” ethical standards ignores the tragic necessity of moral compromise in governance. Consider Abraham Lincoln, widely regarded as one of America’s most effective presidents. During the Civil War, Lincoln suspended habeas corpus—a fundamental constitutional right—and authorized military tribunals for civilians. By the strictest ethical yardsticks, these actions were morally troubling. Yet historians broadly agree that these measures were instrumental in preserving the Union and ultimately abolishing slavery. If Lincoln had insisted on maintaining the “highest” ethical standards at every moment, he might have lost the war and, paradoxically, failed to achieve the profoundly moral objective of emancipation. Effectiveness sometimes demands actions that fall short of moral perfection.
Furthermore, the most effective public officials are often those who can make difficult, pragmatic decisions rather than those who dogmatically adhere to abstract moral codes. During the 2008 financial crisis, the U.S. government authorized massive bailouts of failing banks—institutions whose reckless behavior had precipitated the crisis. From a purely moral standpoint, allowing these institutions to collapse might have seemed just. However, Treasury Secretary Henry Paulson and Federal Reserve Chair Ben Bernanke, motivated by pragmatic concerns about systemic collapse, orchestrated interventions that, while morally ambiguous, likely averted a second Great Depression. Their effectiveness was measured not by moral purity but by the catastrophic outcomes they prevented.
Admittedly, proponents of the prompt’s position might argue that without ethical standards, leaders become corrupt and lose public trust—a fair concern. History provides ample evidence of leaders whose moral failings destroyed their effectiveness: Richard Nixon’s deception during Watergate ultimately paralyzed his presidency. However, this objection conflates the absence of ethics with the presence of imperfect ethics. The relevant standard is not whether a leader achieves moral perfection, but whether they maintain sufficient integrity to sustain public confidence while exercising the practical judgment that governance requires. The line between Nixon’s mendacity and Lincoln’s wartime compromises is not a technicality; it is the difference between self-serving corruption and principled pragmatism.
In conclusion, while ethical standards are indispensable to legitimate leadership, demanding the “highest” moral purity misrepresents the realities of governance. The most effective leaders throughout history have been those who possessed enough integrity to maintain trust, coupled with enough flexibility to make the difficult choices that leadership inevitably entails. We would do well to judge our public officials not by their proximity to sainthood, but by the wisdom and courage with which they navigate the moral complexities inherent in the exercise of power.
Issue Topic 2: Technology（科技）
Prompt: “As people rely more and more on technology to solve problems, the ability of humans to think for themselves will surely deteriorate.”
分析題目：
- 核心議題：科技依賴是否必然導致思考能力退化？
- 關鍵詞：“surely”（必然地——太絕對）、“think for themselves”（獨立思考——定義是什麼？）
- 可採立場：不同意——科技重新分配了思考的內容而非消除了思考能力
Brainstorm 雙面觀點：
| 同意方 | 反對方 |
|---|---|
| GPS 讓人失去方向感 | 科技自動化重複性工作，釋放更高層次的思考 |
| 計算機讓心算能力退化 | 程式設計需要強大的邏輯思考能力 |
| 社群媒體演算法塑造我們的觀點 | 網路讓資訊更易取得，有利於獨立研究 |
Model Essay（約 420 words）：
The claim that increasing reliance on technology will “surely” lead to the deterioration of independent human thought is a familiar refrain in cultural criticism, echoing anxieties that have accompanied every major technological advance from the printing press to the internet. While this concern is understandable, I find the argument fundamentally unpersuasive. Technology does not eliminate human thinking; it redirects it toward higher-order cognitive tasks.
The central flaw in the prompt’s reasoning is its narrow definition of “thinking for oneself.” When a calculator performs arithmetic, it does indeed relieve us of the need to compute manually. But this does not mean thinking has deteriorated—it means that our cognitive resources have been liberated for more complex reasoning. A civil engineer designing a bridge does not spend mental energy on multiplication tables; instead, she applies sophisticated understanding of physics, materials science, and environmental factors to solve problems that no amount of unaided arithmetic could address. The calculator is not a substitute for thinking; it is a tool that enables thinking at a higher level of abstraction.
Moreover, the very technologies that the prompt frames as threats to cognition often demand new and sophisticated forms of mental engagement. Consider programming: software development requires systematic logical reasoning, creative problem-solving, and the ability to anticipate edge cases and failure modes. Far from atrophy, the skills cultivated by engaging with technology represent some of the most demanding cognitive activities humans have ever undertaken. A programmer debugging a complex distributed system is exercising intellectual muscles that a medieval scribe—however skilled—never had the opportunity to develop. Technology has expanded, not contracted, the repertoire of human thought.
Admittedly, there are instances where technology appears to erode specific cognitive capacities. Studies suggest that heavy GPS users show diminished spatial navigation skills, and the ubiquity of search engines may reduce our motivation to commit facts to memory. These are legitimate observations. However, they describe localized skill shifts rather than a general deterioration of independent thought. The human brain has always offloaded cognitive tasks: writing externalized memory, books externalized knowledge, and now digital tools externalize computation and information retrieval. At each stage, the skills rendered less necessary were replaced by capacities that the new medium enabled. Socrates worried that writing would destroy memory; instead, it enabled philosophy.
In conclusion, the fear that technology inevitably erodes independent thinking mistakes the relocation of cognitive effort for its disappearance. While specific skills may atrophy through disuse, the broader arc of technological development has consistently raised the ceiling of what human minds can achieve. The challenge is not to resist technological tools, but to ensure that we use them as platforms for higher thought rather than excuses for mental passivity.
Issue Topic 3: Government / Society（政府與社會）
Prompt: “Governments should place few, if any, restrictions on scientific research and development.”
分析題目：
- 核心議題：科學研究該不該被政府限制？
- 關鍵詞：“few, if any”（幾乎不該有限制——極端立場）
- 可採立場：部分同意——基礎科學應自由，但某些領域（基因編輯、AI 武器）需要節制
Model Essay（約 430 words）：
The proposition that governments should impose “few, if any” restrictions on scientific research speaks to an Enlightenment ideal: that the unfettered pursuit of knowledge is an unqualified good. While I share the instinctive sympathy for scientific freedom that underpins this claim, I believe it overlooks the genuine ethical hazards that certain lines of inquiry present. A more nuanced position is warranted: governments should indeed minimize restrictions on most research, but must maintain the capacity for oversight in areas where the potential for catastrophic harm is demonstrable.
The strongest argument for minimal government interference in science is historical: attempts to suppress research have almost invariably been retrograde and futile. When the Catholic Church forced Galileo to recant his heliocentric views, it delayed the acceptance of astronomical truth but could not ultimately prevent it. When the Soviet Union elevated Lysenko’s pseudoscientific genetics over Mendelian inheritance to align with Marxist ideology, it crippled Soviet biology for decades. These episodes illustrate that political interference in science, driven by ideology rather than evidence, tends to produce ignorance rather than safety. For the vast majority of scientific inquiry—from particle physics to botanical taxonomy—government should adopt a posture of benign non-interference.
However, the “few, if any” formulation collapses when confronted with research that carries existential or severe ethical risks. Consider gain-of-function research on pathogens: studies that deliberately enhance the transmissibility or lethality of viruses like H5N1 avian influenza. Proponents argue that such work helps us prepare for natural pandemics; critics counter that a laboratory accident could unleash a catastrophe far worse than any natural outbreak. When the stakes involve pandemics that could kill millions, the precautionary principle demands regulatory oversight. Similarly, the development of lethal autonomous weapons—“killer robots” that can select and engage targets without human intervention—raises profound moral questions that cannot be left solely to engineers and military contractors. Governments have a legitimate role in establishing ethical boundaries when research carries risks that markets and laboratories are ill-equipped to evaluate.
Crucially, the distinction I am drawing is not between “safe” and “dangerous” science broadly, but between research whose primary risks are epistemic (producing false knowledge, which the scientific method self-corrects) and research whose risks are physical and potentially irreversible. The scientific community has robust mechanisms for debunking bad theories; it has no equivalent mechanism for containing an engineered pandemic or recalling autonomous weapons once they proliferate. Government regulation should be narrowly tailored to this latter category, employing transparent, evidence-based frameworks rather than political or ideological litmus tests.
In conclusion, while the spirit of the prompt—defending scientific inquiry from political encroachment—is worthy of endorsement, its absolutism is untenable. The optimal policy framework maximizes research freedom as a default while maintaining carefully bounded regulatory capacity for the narrow domain of inquiry where the downside risks are genuinely catastrophic. Science needs liberty, but liberty without limits at the precipice is not enlightenment—it is recklessness.
Issue Topic 4: Education & Learning（教育與學習）
Prompt: “Competition for high grades seriously limits the quality of learning at all levels of education.”
分析題目：
- 核心議題：追求高分是否限制了真正的學習品質？
- 關鍵詞：“seriously limits”（嚴重限制）、“at all levels of education”（所有教育層級——太絕對）
- 可採立場：部分同意——過度競爭有問題，但適度競爭在某些層級有激勵作用
Model Essay（約 400 words）：
The assertion that grade competition “seriously limits” learning quality across all educational levels resonates with a growing chorus of educational reformers who argue that extrinsic motivators undermine intrinsic curiosity. I find this argument largely, though not entirely, convincing. While excessive focus on grades can indeed distort learning, a blanket condemnation of competition overlooks contexts in which it serves a constructive function.
The most compelling evidence for the prompt’s position comes from research on intrinsic motivation. Psychologists Edward Deci and Richard Ryan’s self-determination theory demonstrates that when external rewards—such as grades—become the primary focus, individuals experience diminished intrinsic motivation and engage in more superficial cognitive processing. A student who reads Shakespeare solely to identify themes likely to appear on an exam processes the text fundamentally differently from one who reads for aesthetic and intellectual engagement. The former skims for extractable answers; the latter wrestles with ambiguity. This phenomenon, documented across decades of educational psychology research, supports the claim that grade-centric education encourages strategic compliance rather than deep understanding.
Furthermore, the competition paradigm structurally disadvantages certain forms of learning that are inherently resistant to standardized measurement. Creativity, critical thinking, collaboration, and intellectual risk-taking are notoriously difficult to grade and are often penalized in competitive environments. A student who pursues an unconventional thesis risks a lower grade than one who competently executes a safe, predictable argument. Over time, this dynamic trains students to prioritize rhetorical polish over intellectual substance—a distortion of what education at its best should cultivate.
However, the prompt overreaches in asserting that this problem applies “at all levels of education.” In certain professional and graduate contexts, competitive assessment serves a legitimate gatekeeping function. Medical licensing examinations, for instance, are fundamentally competitive assessments that ensure public safety; a surgeon who scraped by with minimal competence poses a genuine threat. Moreover, for some students, particularly in later educational stages, competition provides motivational structure that purely intrinsic approaches may not. The key variable is not whether competition exists but whether the assessment method authentically measures the competencies it purports to evaluate.
In conclusion, while I share the prompt’s concern that obsessive grade competition impoverishes learning, the solution is not to eliminate competition altogether but to redesign assessment systems so that the behaviors rewarded by grades align with genuine intellectual development. The goal should be to make competition serve learning, rather than allowing it to substitute for it.
Issue Topic 5: Ethics & Values（倫理與價值觀）
Prompt: “The best way to understand a society is to examine the character of the men and women that the society chooses as its heroes or its role models.”
分析題目：
- 核心議題：一個社會崇拜的英雄是否能代表該社會的真實面貌？
- 關鍵詞：“best way”（最好的方式——太絕對，有更好的方式嗎？）
- 可採立場：部分同意——英雄確實反映社會價值，但不是「最好的」方式
Model Essay（約 410 words）：
The claim that examining a society’s chosen heroes constitutes the “best way” to understand that society is an intriguing proposition with considerable intuitive appeal. A culture’s heroic ideals do indeed function as a mirror reflecting its deepest aspirations. However, I find the superlative “best” difficult to defend: heroes reveal what a society wishes to be, not necessarily what it actually is, and other analytical lenses offer equally indispensable insights.
The strength of the prompt’s position lies in the fact that heroes are never chosen arbitrarily; they embody values that a particular culture elevates above others. Ancient Athens celebrated Odysseus—cunning, resourceful, and silver-tongued—reflecting a society that prized cleverness and rhetorical skill. By contrast, Sparta lionized Leonidas, the warrior-king who died with his 300 soldiers at Thermopylae, embodying martial valor and sacrificial duty. These contrasting heroic archetypes genuinely illuminate deep differences between Athenian and Spartan civilizations. Similarly, the transformation of American cultural heroes from frontier individualists like Daniel Boone to figures like Martin Luther King Jr. traces a meaningful arc in the nation’s evolving moral imagination. There is real interpretive power in this approach.
Nevertheless, heroes reveal aspirational values, not lived realities—and the gap between the two can be vast. The Soviet Union celebrated the Stakhanovite worker as its heroic ideal: a coal miner who allegedly extracted 102 tons of coal in a single shift, exceeding his quota by 1,400 percent. This heroic narrative spoke to official ideology about labor and productivity, but it told one almost nothing about the actual conditions of Soviet life: the shortages, the fear, the cynicism with which most citizens regarded propaganda. A society’s chosen heroes can be instruments of mystification as easily as vehicles of truth. To understand the Soviet Union, one would learn far more from examining its prison archives than its heroic statuary.
Moreover, the “best” way to understand a society may depend on what aspect of that society one wishes to understand. If one seeks insight into a culture’s professed values and collective self-image, its heroes are indeed an excellent object of study. But if one wishes to understand social stratification, economic organization, or political power dynamics—the structures that shape daily life far more than heroic narratives—then analyzing institutions, laws, and economic data would yield richer understanding. The daily experience of a minimum-wage worker in contemporary America is shaped far more profoundly by labor law and market forces than by the nation’s reverence for figures like Steve Jobs or Rosa Parks.
In conclusion, while a society’s pantheon of heroes offers a valuable and revealing window into its collective psyche, it is neither the sole nor the “best” lens through which to study a civilization. A comprehensive understanding requires triangulating between a culture’s ideals, its institutions, and the lived experiences of its people. Heroes tell us what a society dreams of becoming; the gap between those dreams and daily reality is where the deepest understanding resides.
三、Argument Essay（論證分析）
3.1 什麼是 Argument Essay？
Argument Essay 是 GRE 寫作中最容易被誤解的題型，也是台灣學生失分最嚴重的項目。
關鍵理解：Argument Essay 不是要你表達立場！你要做的事情是：分析一段論證中的邏輯漏洞。
題目會給你一段短文（通常來自模擬的報紙社論、公司備忘錄、政府報告等），你要：
- 找出論證中的邏輯缺陷（logical fallacies）
- 解釋為什麼這些缺陷會削弱論證的說服力
- 提出什麼樣的額外證據可以加強或驗證這個論證
千萬不要做的：
- 不要說你同意或不同意原文的結論
- 不要提出你自己的解決方案
- 不要討論題目沒有給的資訊當作事實
常見錯誤
台灣學生最大的錯誤就是在 Argument Essay 中寫 “I disagree with the author’s conclusion because…” ——這樣寫直接離題！Argument Essay 考的是你的邏輯分析能力，不是你的立場。請把每句話的主語都放在「論證的邏輯」上，而不是「你的意見」上。
3.2 Issue vs. Argument 對比表
| 面向 | Issue Essay | Argument Essay |
|---|---|---|
| 任務 | 針對一個議題表達並捍衛你的立場 | 分析一段論證中的邏輯漏洞 |
| 你需要做 | 提出理由＋例子支持你的觀點 | 指出邏輯缺陷並解釋為何它們削弱論證 |
| 你不能做 | — | 表達你是否同意結論、提出自己觀點 |
| 寫作風格 | 辯論式、說服式 | 分析式、批判式 |
| 例子 | 需要自己的例子 | 不需要外部例子，只需分析原文邏輯 |
| 開頭句 | ”I agree/disagree that…" | "The argument that… is flawed for several reasons.” |
| 結尾 | 重申立場＋總結 | 總結邏輯缺陷＋建議需要什麼證據 |
3.3 Argument Essay 結構模板
Paragraph 1: Introduction
- 簡述原文論證的結論和主要證據
- 指出論證有 N 個重大缺陷
- (不要說你同意或不同意)
Paragraph 2-4: Body（每個段落分析一個邏輯漏洞）
每段結構：
1. 指出一個具體的邏輯漏洞
2. 解釋為什麼這是一個問題
3. 舉例說明其他可能的解釋
4. 提出需要什麼額外證據
Paragraph 5: Conclusion
- 總結主要邏輯缺陷
- 不是說結論錯了，而是說論證不充分
- 建議額外證據可能如何改變結論
Argument Essay 句型框架
Introduction:
The argument that [簡述結論] rests on a chain of reasoning that, upon examination, reveals several significant logical vulnerabilities. While the conclusion may appear plausible at first glance, the evidence provided does not adequately justify it. The argument's weaknesses fall into [數字] major categories: [簡述各類別].
OR:
In [publication/source], the author concludes that [簡述結論], citing [簡述證據]. Although this line of reasoning may seem compelling initially, closer scrutiny reveals that it relies on unwarranted assumptions and insufficient evidence. Before accepting the recommendation, one would need to address several critical gaps in the argument's logic.Body Paragraph（每個邏輯漏洞）：
First, the argument makes an unwarranted leap by assuming that [指出因果跳躍].
The author appears to reason that because [A 發生], therefore [B 必然發生].
However, this causal link is far from established.
It is entirely possible that [提出替代解釋].
For instance, [具體情境說明替代可能性].
Without evidence that [需要的證據], this assumption remains speculative.
Furthermore, the argument relies on [描述第二個問題 — 例如類比不當、數據偏誤等].
The author cites [原文的數據或例子] as support for [原文的推論], yet this evidence is problematic because [解釋為什麼].
The [數據/調查/例子] may not be representative because [提出問題].
To strengthen this line of reasoning, the author would need to provide [需要的額外資訊].Conclusion:
In sum, while the conclusion that [原文結論] is not necessarily incorrect, the argument advanced to support it is riddled with logical gaps. The author's reliance on [主要問題一] and [主要問題二], combined with the failure to account for [替代解釋], undermines the persuasiveness of the reasoning. A more rigorous case would require [列舉需要的證據類型]. Until such evidence is provided, the argument remains unconvincing, and the proposed course of action should be approached with considerable caution.3.4 八大常見邏輯謬誤（Logical Fallacies）
Fallacy 1: Post Hoc Ergo Propter Hoc（錯誤歸因／因果倒置）
定義： 因為 A 發生在 B 之前，就斷定 A 是 B 的原因。
GRE 常見情境：
- 「我們實施了新政策後，銷售額上升了 → 所以新政策導致銷售額上升」
- 「市長上任後犯罪率下降了 → 所以市長的政策導致犯罪率下降」
分析句型：
The argument commits the post hoc fallacy by assuming that because [A] preceded [B], [A] must have caused [B]. Temporal sequence alone does not establish causation. The improvement could be attributable to [替代原因], or may simply reflect coincidence. Without controlling for confounding variables, this causal inference is unjustified.
Fallacy 2: False Analogy（不當類比）
定義： 拿兩個不夠相似的事物來類比，假設它們在其他方面也會相似。
GRE 常見情境：
- 「A 市實施這政策成功了，所以我們 B 市也應該實施」
- 「這家公司在軟體業的策略有效，所以他們在家電業也會成功」
分析句型：
The argument draws an analogy between [X] and [Y], assuming that what worked in one context will work in the other. However, the analogy is undermined by potentially significant differences, such as [差異一] and [差異二]. Unless the author can demonstrate that these contexts are sufficiently similar in all relevant respects, the analogy provides little persuasive force.
Fallacy 3: Hasty Generalization（過度概括／以偏概全）
定義： 用不充分或非代表性的樣本得出普遍結論。
GRE 常見情境：
- 「我們訪談了 10 位顧客，他們都說滿意 → 所以所有顧客都滿意」
- 「這所學校的前三名畢業生表現很好 → 所以這學校的教育品質很高」
分析句型：
The argument relies on a sample that may not be representative of the broader population about which it draws conclusions. [描述樣本問題：太小、自我篩選、時間範圍太短等]. The results could be skewed by selection bias or random variation. A larger, more systematically collected sample would be necessary before generalizing.
Fallacy 4: False Dichotomy（假兩難／非黑即白）
定義： 把複雜情況簡化為只有兩個選項，忽略其他可能性。
GRE 常見情境：
- 「要嘛我們砍預算，要嘛公司破產」
- 「大學應該只重視 STEM 或只重視人文」
分析句型：
The argument presents a false dichotomy, framing the situation as a choice between only [A] and [B]. This binary framing ignores a spectrum of intermediate or alternative approaches, such as [例子]. By artificially restricting the range of options, the argument prematurely forecloses solutions that might reconcile competing concerns.
Fallacy 5: Unsubstantiated Assumption（未證實的假設）
定義： 論證的關鍵前提並未被證明或提供證據。
GRE 常見情境：
- 「因為學生喜歡線上課程 → 線上課程的學習效果更好」（假設：喜歡 = 有效）
- 「調查顯示更多人選擇國產車 → 國產車品質提升了」（假設：選擇原因 = 品質）
分析句型：
The argument rests on the unsubstantiated assumption that [描述假設]. The author provides no evidence to support this crucial premise. Alternative explanations, such as [替代解釋], could account for the observed phenomenon equally well. Until the author substantiates this foundational claim with empirical evidence, the entire chain of reasoning remains precarious.
Fallacy 6: Correlation vs. Causation（相關不等於因果）
定義： 兩件事同時發生或同步變化，就斷定一者是另一者的原因。
GRE 常見情境：
- 「冰淇淋銷量和溺水人數同時上升 → 吃冰淇淋導致溺水」（其實都是因為夏天）
- 「使用社群媒體時間越長，憂鬱症比例越高 → 社群媒體導致憂鬱」（可能是反向因果或其他變數）
分析句型：
The argument conflates correlation with causation. The fact that [A] and [B] are associated does not establish that [A] causes [B]. It is equally plausible that [B] causes [A], that a third variable [C] drives both, or that the correlation is entirely coincidental. Establishing causality would require controlled experimental or longitudinal data, not merely observational patterns.
Fallacy 7: Straw Man（稻草人謬誤）
定義： 把對方的立場扭曲成一個比較弱的版本，然後攻擊這個扭曲的版本。
GRE 常見情境：
- 主張適度環保法規 → 被說成「想要摧毀所有產業」
- 主張教育改革 → 被說成「想要廢除所有考試」
分析句型：
The argument misrepresents the opposing position by reducing it to a caricature—a classic straw man fallacy. The actual position, which may involve nuance such as [描述真正的立場], is far more defensible than the extreme version the author attacks. By dismantling an argument that no reasonable proponent actually holds, the author avoids engaging with the real substance of the debate.
Fallacy 8: Either-Or with Insufficient Evidence（證據不足以支持結論）
定義： 即使論證中的所有前提都成立，結論也不必然成立；證據和結論之間有邏輯跳躍。
GRE 常見情境：
- 「我們的競爭對手增加了廣告預算 → 我們也該增加」（沒證明增加預算有效）
- 「學生要求更多選修課 → 所以必修課制度應該廢除」（跳太大）
分析句型：
Even if one were to grant all of the author’s factual claims, the conclusion does not logically follow. The argument suffers from a non sequitur: the evidence presented does not uniquely point to the recommended course of action. The same set of facts could support entirely different conclusions, such as [替代結論]. The author fails to explain why the preferred interpretation is more warranted than others.
3.5 實戰 Argument 練習（3 題）
Argument Practice 1: Business / Marketing
Prompt: The following appeared in a memo from the vice president of marketing at Dura-Socks, Inc.
“A recent study of Dura-Socks customers suggests that our company is wasting the money it spends on its current advertising campaign. In the study, researchers asked Dura-Socks customers whether they recalled seeing any Dura-Socks advertisements in the past 60 days. Only 35 percent of customers reported recalling such advertisements. Clearly, 65 percent of our customers are not being reached by our marketing efforts. If we want to continue to grow as a company, we must completely overhaul our advertising strategy and redirect our marketing budget toward social media platforms, which studies show are the most effective channels for reaching consumers in our target demographic.”
Step 1: 找出邏輯漏洞
- Recall vs. Reach（記憶不等於未接觸）： 沒有回憶起廣告不等於沒有看到廣告。消費者每天接觸數千則廣告，大多數被潛意識處理而非明確記憶。
- Causation Assumption（因果假設）： 低回憶率不一定代表廣告無效——廣告可能透過潛意識曝光（mere exposure effect）影響購買行為。
- Binary Solution（假兩難）： 完全翻修策略 vs. 維持現狀不是唯二選擇。可以調整現有策略而非完全廢棄。
- Social Media Assumption（社群媒體假設）： 沒證明社群媒體對這個特定產品和族群有效。襪子消費者的媒體使用習慣可能與一般族群不同。
- Sample Validity（樣本問題）： 只調查「現有顧客」——如果廣告的目的是獲取新客戶，這個調查無法反映廣告的獲客效果。
Model Essay（約 450 words）：
The vice president of Dura-Socks argues that the company’s advertising campaign is wasteful and should be completely overhauled in favor of social media marketing. This recommendation is based on a customer survey in which only 35 percent of respondents recalled seeing Dura-Socks advertisements. While the memo raises legitimate questions about marketing effectiveness, its reasoning contains several logical gaps that render its prescription premature at best.
The most fundamental problem with the argument is its conflation of advertisement recall with advertisement effectiveness. The memo assumes that customers who do not consciously remember seeing an ad were not reached by it. Modern advertising research, however, has thoroughly documented the “mere exposure effect”—the phenomenon whereby repeated, even subconscious, exposure to a brand increases familiarity and preference without producing explicit memories. A customer might choose Dura-Socks at the store not because she remembers a specific commercial, but because the brand “feels familiar.” The 65 percent who did not recall ads may nonetheless have been influenced by those ads. The survey, as described, measures conscious recall rather than behavioral impact, making it an inadequate tool for assessing the campaign’s true effectiveness.
Furthermore, even if we were to accept that the current campaign has limited reach, the argument leaps to the conclusion that social media marketing is the appropriate alternative without providing evidence specific to Dura-Socks’ circumstances. The memo cites generic “studies” showing social media’s effectiveness for reaching “consumers in our target demographic,” yet it fails to establish that these studies apply to the sock market specifically. The purchasing behavior of sock consumers may differ substantially from the general population: socks are often bought in physical stores as impulse purchases or routine replacements, and the demographic that buys premium socks may not be the demographic most engaged on social media platforms. The author assumes, without justification, that a strategy effective for other products or demographics will translate to Dura-Socks’ particular market position.
Additionally, the argument presents a false dichotomy between maintaining the current campaign and “completely overhauling” it. This binary framing ignores a range of intermediate strategies that might be more prudent. The company could, for instance, conduct A/B testing by shifting a modest portion of its budget to social media while maintaining its existing channels, then comparing results. It could refine its current campaign’s creative elements or targeting parameters rather than discarding the entire approach. By insisting on a wholesale transformation, the author forecloses incremental, evidence-based optimization—which is how most successful marketing strategies actually evolve.
Finally, the survey methodology raises significant concerns. By sampling only existing customers, the study systematically excludes the very population that advertising is often designed to reach: prospective customers who have not yet purchased the product. If the advertising campaign’s primary objective is customer acquisition, a survey of current customers tells us little about whether the ads are reaching and converting new buyers. The sample may be further biased if the survey attracted respondents who are particularly engaged with the brand or who have strong opinions—either positive or negative.
In conclusion, while the vice president’s concern about marketing ROI is understandable, the argument for a complete strategic overhaul is built on a foundation of logical vulnerabilities. A more persuasive case would require evidence that advertisement recall correlates with purchasing behavior for Dura-Socks specifically, data on social media marketing effectiveness within the sock industry, and a more rigorous survey methodology that captures the perspectives of both current and prospective customers. Until such evidence is provided, the company would be wise to pursue a more measured, experimental approach to revising its marketing strategy.
Argument Practice 2: Education
Prompt: The following appeared in a letter to the editor of a university newspaper.
“Last year, the university allocated 5 million expansion of the fitness center to include Olympic swimming pools and additional sports courts.”
Step 1: 找出邏輯漏洞
- Post Hoc / Causation（因果倒置）： 健身中心翻新後壓力報告減少、GPA 上升，不代表是翻新造成的。可能有其他原因（新的諮商服務、較輕鬆的課程安排等）。
- Reporting Bias（通報偏差）： 壓力報告減少 ≠ 壓力減少。學生可能因為各種原因不願意去諮商中心通報。
- Correlation Gap（相關性跳躍）： 沒有證據顯示使用健身中心的學生就是壓力減少或 GPA 提升的那群學生。
- Extrapolation（外推問題）： 就算 $2M 翻新有效，$$5M 擴建也未必會有比例性的效果。邊際效益遞減。
- GPA 變化解釋： 0.1 的 GPA 變化可能出於評分標準改變、課程難度調整、或是隨機波動。
Model Essay（約 440 words）：
The letter to the editor argues that the university’s investment in fitness center renovations caused improvements in student mental health and academic performance, and that therefore a further $5 million expansion should be approved. While the observed correlations are suggestive, the argument’s causal reasoning suffers from multiple logical defects that render its policy recommendation unjustified.
The argument’s central flaw is a classic post hoc fallacy. The author observes that decreased stress reports and increased GPA followed the fitness center renovation, and concludes that the renovation caused these improvements. However, temporal sequence alone offers no proof of causation. Any number of concurrent changes could account for the observed trends. The university may have expanded its counseling services, adjusted academic policies, or hired more accessible faculty members. Broader societal factors—such as a recovering economy reducing financial stress among students—could also be at work. Without controlling for these and other potential variables, attributing the improvements specifically to the fitness center is speculative at best.
Furthermore, the assumption that a decrease in counseling center visits indicates improved mental health is problematic. Students might be reporting less stress to the counseling center not because they are less stressed, but because they are seeking support elsewhere, self-medicating, or simply reluctant to seek help due to stigma. Paradoxically, an effective mental health awareness campaign might increase counseling center visits by encouraging help-seeking behavior. The metric cited—number of reports to the counseling center—reflects service utilization, not the underlying prevalence of mental health issues. A campus-wide anonymous survey would provide far more reliable data about student wellbeing than the counseling center’s caseload.
Even granting the mental health improvement, the argument provides no evidence linking fitness center usage to the students whose GPAs improved. The aggregate rise from 3.1 to 3.2 could reflect grade inflation, easier course offerings, changes in the student population, or any combination of factors unrelated to physical exercise. To establish the claimed relationship, the author would need to demonstrate that students who used the renovated fitness center experienced greater GPA improvements than those who did not, while controlling for variables such as course load and prior academic performance. In the absence of such disaggregated data, the connection between recreation and academic performance remains conjectural.
Perhaps most critically, even if the 5 million investment would yield proportional or even positive returns. The principle of diminishing marginal returns suggests that the initial investment—addressing the most glaring deficiencies in recreational facilities—likely produced the greatest benefit. Adding Olympic swimming pools may serve a small subset of competitive swimmers while offering negligible value to the vast majority of students. The author would need to conduct a cost-benefit analysis comparing this $5 million expenditure against alternative uses of the funds, such as expanding mental health services directly, hiring additional faculty, or increasing financial aid.
In conclusion, the letter weaves an appealing narrative of cause and effect that crumbles under logical scrutiny. The correlation between fitness center renovation and positive campus outcomes does not, by itself, establish that the renovation caused those outcomes, let alone that additional investment is warranted. Before the university commits millions of dollars to further expansion, it should gather rigorous evidence on facility usage patterns, conduct controlled assessments of student wellbeing, and evaluate whether recreational spending represents the most effective use of limited resources.
Argument Practice 3: Public Policy
Prompt: The following is a recommendation from the board of directors of the City of Mapleton Transportation Department.
“Three years ago, the neighboring city of Oakville implemented a congestion pricing system, charging drivers a fee to enter the city center during peak traffic hours. Since then, Oakville has seen a 25 percent reduction in traffic volume in the downtown area and a 10 percent increase in public transit ridership. Meanwhile, Mapleton’s downtown traffic has worsened over the same period, with average commute times increasing by 20 percent. A congestion pricing system modeled on Oakville’s would produce similar benefits for Mapleton. In addition, the revenue generated from congestion fees could fund the expansion of Mapleton’s bus network, further improving transportation options for our residents.”
Step 1: 找出邏輯漏洞
- False Analogy（不當類比）： Oakville 和 Mapleton 可能在地理、人口、公共交通基礎建設上有重大差異。
- Alternative Causes for Oakville（替代解釋）： Oakville 的改善可能來自其他因素（新增自行車道、油價上漲、遠距工作增加等），不一定是 congestion pricing 的功勞。
- Conflicting Trends（趨勢混淆）： Mapleton 惡化期間 Oakville 改善——這不代表兩者有因果關係。Mapleton 可能面臨不同的人口成長或經濟狀況。
- Revenue Assumption（收入假設）： 預設 congestion pricing 會產生足夠收入來擴建公車網絡——但如果政策太成功（大家都不開車進城），收入反而會下降。
- Implementation Feasibility（實施可行性）： 沒有討論政治可行性、對低收入駕駛的影響、或是 Mapleton 是否已有替代交通基建（公車、腳踏車）。
Model Essay（約 430 words）：
The Mapleton Transportation Department’s recommendation that the city should adopt Oakville’s congestion pricing model relies on an appealing but ultimately insufficient comparison between the two cities. While the proposal’s objectives—reducing traffic and funding transit improvements—are laudable, the argument’s logical architecture contains several critical weaknesses that undermine its persuasiveness.
The argument’s foundational move is an analogy: if congestion pricing worked in Oakville, it will work in Mapleton. However, the author provides no information about whether the two cities are comparable in the ways that matter for transportation policy. Congestion pricing succeeds when drivers have viable alternatives, such as robust public transit or bike-friendly infrastructure. If Oakville has an extensive subway or light rail system and Mapleton relies primarily on a limited bus network, the policy’s effects could differ dramatically. Mapleton residents with no practical alternative to driving might simply absorb the fee, generating revenue but failing to reduce traffic—or worse, they might be disproportionately burdened by costs without receiving commensurate benefits. A meaningful comparison would require detailed analysis of each city’s transportation infrastructure, commuter demographics, and geographic layout.
Furthermore, the argument attributes Oakville’s improvements entirely to congestion pricing while ignoring other potential explanatory factors. The three-year period in question may have coincided with an economic downturn that reduced commuting overall, a rise in remote work arrangements, or gas price increases that independently discouraged driving. The 10 percent increase in public transit ridership might reflect service improvements unrelated to congestion pricing, such as expanded routes, reduced fares, or new train lines. Without isolating the effect of the congestion fee from these confounding variables, the claim that congestion pricing “produced” these benefits remains unverified. Rigorous policy evaluation would require comparing Oakville’s outcomes against a control city with similar characteristics that did not implement congestion pricing.
The recommendation also makes an unwarranted assumption about the relationship between Mapleton’s worsening traffic and the absence of congestion pricing. The fact that Mapleton’s commute times increased while Oakville’s decreased does not necessarily imply that adopting Oakville’s policy would reverse Mapleton’s trend. Mapleton may be experiencing population growth, commercial development, or road construction projects that are fundamentally distinct from the conditions Oakville faced. The policy that mitigated moderate congestion in one context may be wholly inadequate for the more severe or structurally different conditions in another.
Finally, the argument’s revenue projection is internally inconsistent. If the congestion fee succeeds in dramatically reducing downtown driving, the revenue base shrinks proportionally. Cities that have implemented congestion pricing—London, Stockholm, Singapore—have found that revenue forecasting is notoriously difficult precisely because the policy’s behavioral effects are hard to predict. Mapleton cannot responsibly plan a bus network expansion around an uncertain revenue stream. A more prudent approach would involve pilot testing, phased implementation, and contingency planning rather than an all-or-nothing commitment predicated on optimistic projections.
In conclusion, the transportation department’s recommendation extrapolates far beyond what the evidence can support. The analogy between Oakville and Mapleton is unearned, the causal attribution is unsubstantiated, and the financial assumptions are unjustified. Before proceeding, the board should commission a feasibility study comparing the two cities across relevant dimensions, gather evidence isolating the specific impact of congestion pricing in Oakville, and develop revenue projections based on conservative, empirically grounded assumptions. Urban transportation policy deserves a more rigorous evidentiary foundation than the one this argument provides.
四、台灣學生寫作進階技巧
4.1 冠詞（Articles）使用
冠詞是台灣學生寫作中最常出錯的項目。GRE 評分雖然不苛求完美文法，但系統性的冠詞錯誤會拉低語言分數。
| 規則 | 說明 | 例子 |
|---|---|---|
| a/an | 首次提到、非特定、可數單數 | ”I have a proposal for solving the problem.” |
| the | 特定對象、唯一、前文提過、最高級 | ”The solution proposed earlier has merit.” |
| 無冠詞 | 複數泛指、不可數泛指、專有名詞 | ”Government should protect citizens.” (泛指政府) |
威威老師小提醒
寫完每段後，快速掃一遍所有的名詞：如果是單數可數名詞，前面有沒有冠詞？如果名詞前面有形容詞，冠詞是加在形容詞前面的：an extremely important issue。
4.2 平行結構（Parallelism）
GRE 高分作文的一個顯著特徵是平行結構的使用。這不只是美觀問題——平行結構提升文章的可讀性和節奏感。
不好的寫法：
The policy would reduce traffic, improving air quality, and to save money for commuters.
好的寫法：
The policy would reduce traffic, improve air quality, and save money for commuters.
更多範例：
| 結構 | 範例 |
|---|---|
| not only…but also | ”This approach not only respects individual autonomy but also promotes social welfare.” |
| either…or | ”Reformers must either build consensus or accept gridlock.” |
| neither…nor | ”The argument provides neither sufficient evidence nor sound reasoning.” |
| 列舉 | ”The proposal fails because it is expensive, impractical, and inequitable.” |
4.3 避免冗詞與空洞表達
GRE 寫作的時間壓力下，很多學生會用「填充詞」來湊字數。但這些詞不但不加分，還讓你的文章顯得薄弱。
| 避免使用 | 改為 | 原因 |
|---|---|---|
| In my personal opinion, I think that… | — (直接開始寫) | 你的文章就是你的 opinion，不用宣告 |
| It is a well-known fact that… | — (直接寫事實) | 如果是事實，不需要說「這是眾所周知的」 |
| Needless to say… | — | 既然 needless to say，就不要 say |
| Due to the fact that… | because | 更簡潔 |
| In the event that… | if | 更簡潔 |
| A large number of… | many | 更簡潔 |
| In this day and age… | today / currently | 更簡潔 |
| Each and every… | each / every | 選一個就好 |
4.4 提升用詞精準度：常用 GRE 作文動詞替代表
| 基本詞 | GRE 高分替換 | 例句 |
|---|---|---|
| show | demonstrate, illustrate, reveal, indicate | ”This example illustrates the broader principle.” |
| cause | engender, precipitate, give rise to | ”Such policies may engender unintended consequences.” |
| reduce | mitigate, alleviate, diminish, curtail | ”The reform would mitigate the worst effects.” |
| increase | augment, amplify, bolster, enhance | ”The evidence bolsters the author’s central claim.” |
| say/argue | contend, posit, assert, maintain, aver | ”The author contends that the correlation proves causation.” |
| disagree | refute, rebut, challenge, undermine | ”This evidence undermines the argument’s premise.” |
| emphasize | underscore, highlight, accentuate | ”The case underscores the need for caution.” |
| weaken | undermine, erode, compromise | ”The logical gap compromises the argument’s credibility.” |
| support | substantiate, corroborate, buttress | ”Additional data would corroborate the claim.” |
| examine | scrutinize, interrogate, probe | ”The essay scrutinizes the assumptions underlying the proposal.” |
威威老師小提醒
不要每句話都換一個高級動詞！過度使用會顯得矯揉造作。每段用 2–3 個精準的進階詞，搭配自然流暢的普通詞彙，這樣才像真正的學術寫作。GRE 的 e-rater 會偵測詞彙多樣性，但 human rater 會看整體的 readability。
4.5 常見中式英文修正
| 中式英文 | 問題 | 正確英文 |
|---|---|---|
| ”According to my opinion…” | 這裡不用 according to | ”In my opinion…” 或 “I believe…" |
| "The reason is because…“ | redundant（贅字） | “The reason is that…" |
| "Although…but…” | 中文轉折習慣，英文不需要 | ”Although… / …”（去掉 but） |
| “There have many people…” | 應該用 there are | ”There are many people…" |
| "I very much agree…“ | very much 放錯位置 | ”I strongly agree…” 或 “I agree wholeheartedly…" |
| "On the other hand” 只用一次 | 中文「另一方面」只需一個 | 不要寫 “On one hand…on the other hand…on the other hand…” |
五、考前準備清單
5.1 考前一個月
| 週次 | 任務 |
|---|---|
| 第 1 週 | 熟悉所有 Issue 題庫題目（ETS 官網），將 150+ 題分類成 5-8 大類 |
| 第 2 週 | 練習 Issue：每 2 天寫一篇（計時 30 分鐘），共 3-4 篇 |
| 第 3 週 | 練習 Argument：每天分析 1-2 篇 Argument 題目（不一定要寫全文，重點是抓漏洞） |
| 第 4 週 | 混合練習：每 2 天模擬一次完整 AWA（Issue + Argument 各 30 分鐘），共 3-4 次 |
5.2 考前一天的準備
- 複習你寫過的文章，特別是老師或同學給的修改建議
- 瀏覽一遍 Argument 題庫，確認你能在 1-2 分鐘內找出 3 個邏輯漏洞
- 準備好你的萬用例子庫（至少 5 個）
- 不要熬夜！AWA 是考試第一部分，睡眠不足會直接影響寫作思路
5.3 考試當天注意事項
- 打字速度是硬傷！ 如果平常打字慢（低於 30 WPM），考前一定要練習。一篇 500 字的 Essay 在 30 分鐘內寫完，扣掉思考和檢查時間，實際打字時間可能只有 20 分鐘——那意味著你需要每分鐘至少打 25-30 個字。
- 先打大綱！ 花 3-5 分鐘在螢幕上打大綱（直接打在答題區，之後再擴寫），不要只在腦中想。
- 不要離題！ 每寫完一段，確認這段的核心句和你的 thesis 相關。
- 留時間檢查！ 有時間回頭改一個明顯的文法錯誤，可能影響 0.5 分。
六、建議學習資源
- Official GRE Issue Pool & Argument Pool（ETS 官網免費下載——考前必看！）
- The Elements of Style (Strunk & White) — 英文寫作聖經，薄薄一本講完所有寫作原則
- They Say / I Say: The Moves That Matter in Academic Writing (Graff & Birkenstein) — 學術寫作模板書，對 Issue Essay 特別有幫助
- Manhattan Prep GRE Reading Comprehension & Essays — GRE 寫作策略書
- Grammarly — 練習時貼上去檢查文法問題，但考試時當然不能用！
- TypingClub.com — 免費打字練習網站，英打速度慢的同學必練
威威老師小提醒
GRE 寫作是台灣學生最容易被低估的部分，但也是最能透過「刻意練習」提升的部分。
一篇好的 GRE 作文不需要你的英文像母語者，但需要：
- 思路清晰
- 結構完整
- 論證有據
這些都是可以「學得會」的技能——加油，你絕對可以拿到 4.0+！
威威老師的最後一句話
記住威威心法：
寫作是思考的外在表現——先想清楚，自然就寫得清楚！
3 大成功要素：
- 打字速度 ≥ 30 WPM（不練打字 = 寫不完）
- 5 段結構模板（Intro + 2 Body + Concession + Conclusion）
- 批判思考層次（複雜性 + 具體例子 + 反方思考）
6 個月有計畫地練——AWA 從 3.0 衝到 5.0+！
我們頂尖研究所見！🚂🌍