# Research Paper (RP) ## 架構 ![image](https://hackmd.io/_uploads/B1H55vhX1e.png) * Introduction (I): The main purpose of the Introduction is to provide the rationale for the paper, moving from a general discussion of the topic to the particular question, issue, or hypothesis being investigated. A secondary purpose is to attract interest in the topic‐and hence readers. * Methods (M): The Methods section describes, in various degrees of detail, methodology, materials (or subjects), and procedures. This is the narrowest part of the RP. * Results \(R\): In the Results section, the findings are described, accompanied by variable amounts of commentary. * Discussion (D): The Discussion section gives meaning to and interprets the results in a variety of ways. Authors make a series of "points," at least some of which refer to statements made in the Introduction. ### Frequencies of selected features ![image](https://hackmd.io/_uploads/Bk6XiDh7Jl.png) ## Title 1. The title should indicate the topic of the study. 2. The title should indicate the scope of the study (i.e., neither overstating nor understating its significance). 3. The title should be self‐explanatory (不證自明的) to readers in the chosen area. * 加 colon * Problem: Solution * General: Specific * Topic: Method * Major: Minor ## Abstract * 兩種 approach * Results‐drive: it concentrates on the research findings and what might be concluded from them * RP summary: it provides one‐ or two‐sentence synopses of each section of the paper * 描述 what was done 時使用 [indicative sentence](https://hackmd.io/g93aqi7FRryVBUZr7BrVDQ?view=&stext=14880%3A20%3A0%3A1733834044%3AejjW22) * 描述 main findings 時使用 [informative sentence](https://hackmd.io/g93aqi7FRryVBUZr7BrVDQ?view=&stext=15039%3A21%3A0%3A1733834048%3AEBd6Pc) * 需包含 motivation, key point 及 contribution ## Introduction section ### Create‐A‐Research‐Space (CARS) Model 1. Move 1: Establishing a research territory * by showing that the general research area is important, central, interesting, problematic, or relevant in some way (optional) * by introducing and reviewing items of previous research in the area (obligatory) 2. Move 2: Establishing a **niche** * by indicating a gap in the previous research or by extending previous knowledge in some way (obligatory) * Opening sentences (使用 quasi-negative 的句子) * However, few studies ... * Few investigations ... * Few researchers ... * Few attempts ... * However, little information ... * Little attention ... * Little work ... * Little data ... * Little research ... * 技巧 (上 (stronger) <--> 下 (weaker)) * counter‐claiming (something is wrong) * indicating a gap (something is missing) * raising a question or making an inference (something is unclear) * continuing a tradition (adding something) (Science 及 Engineering 最常見) 3. Move 3: Occupying the niche * by outlining purposes or stating the nature of the present research (obligatory) * by listing research questions or hypotheses (probable in some field) * by announcing principal findings (probable in some field) * by stating the value of the present research (probable in some field) * by indicating the structure of the RP (probable in some field) * 兩種 variants: * Purposive \(P\): The author or authors indicate their main purpose or purposes. * Descriptive (D): The author or authors describe the main feature of their research. * 常用詞 * this * the present reported * here * we * **Metadiscoursal (元話語，引導讀者注意力的 sentence)** 會放在句子 **前** 而不是後 * > :o: *In this paper* we present the results of three experiments. * > :x: We present the results of three experiments *in this paper*. ### Citations 1. Pattern 1: reference to single studies (past tense) 可用 verbs: stated, concluded, claimed > Huang (2007) *investigated* the causes of airport delays. 2. Pattern 2: reference to areas of inquiry (present perfect tense) 可用 verbs: argues that, says, states > The causes of airport delays *have been* widely investigated (Hyon, 2004; Huang, 2007; Martinez et aI., 2010). 3. Pattern 3: reference to state of current knowledge (present tense) > The causes of airport delays are complex (Hyon, 2004; Huang, 2007, Martinez et aI., 2010). :::success 可以多個 pattern 交互著用，增加變化性 ::: ## Method section ### Extended Description and Condensed Description in Method section #### Extended Description * 使用時機 * Providing useful background knowledge (e.g., Through definitions and examples) * Using descriptions of procedural steps, rather than citations and/or acronyms * Including a number of justifications (e.g., To determine this value, we …) * Using cognitive or volitional verbs (e.g., We believed; we wanted to) * Including by + ‐ing + verb + how statements (this was done by reversing the order) * Employing a wide range of linking words and phrases (e.g., Time expressions, such as next or prior to) * > **To detect groups among the specimens and extract the variables that best diagnose the groups**, we *used* principal components analysis (PCA). Before conducting the analysis, we *standardized* all measurements so that each variable would have a mean of 0 and a standard deviation of 1. **For the PCA**, we *included* only continuous characters. **To avoid weighting characters**, we *excluded* characters that are probably genetically redundant, as revealed by high values for the Pearson correlation coefficient between all possible pairs of characters. #### Condensed Description * 使用時機 * Assume readers have relevant background knowledge * Sometimes use citations or acronyms to refer to processes (e.g., A corpus was designed following romer (2010)) * Have few justifications * Use few or no cognitive or volitional verbs * Choose to avoid by + ‐ing + verb + how statements * Employ few linking phrases * > DNA was extracted from tissue and feather samples using the Qiagen DNeasy Extraction kit. An addition of dithiothreitol was used for samples from feathers. Polymerase chain reaction (PCR) was carried out using two primers pairs for cytochrome B (CytB; Sorenson et aI., 1999). PCR and sequencing was done following protocols in Mindell et al. (1997). ### 常用的 Purpose Clause * determine * test * avoid * establish * illustrate * obtain * reduce * understand * to + V * in order to + V * In an effort to evaluate * To further test this hypothesis * To determine the cost * In the interest of obtaining useful data ### 常用的 Cognitive and Volitional Verbs * ... was considered * was thought ... * ... with a number of aims in mind * we wanted ... * we also felt it is prudent ... * we eventually settled on ... ### 常用的 Temporal Link Phrases * During the data collection, * Prior to collecting this information, * On arrival on campus, the participants * In the follow‐up phase of the study, we * After the interview, subjects were * Before conducting the analysis, * After incubating ... ### 常用的 Causal or Connective Phrases * Based on the feedback from the pilot study, * On the basis of the literature review, * Because of privacy issues, we * In spite of these issues, we * In light of these unexpected findings, ## Result section ### Special Verb Agreements 標準的句子的 verb 都是以 subject noun 做 agree，但如果 subject noun 是 **fraction**, **proportion** 或 **percentage** 的話，verb 就要以 **離最近的 noun** 做 agree > A set of 200 questionnaires *was* distributed. > A large proportion of students *are* distracted when they cross a street while talking on the phone. > Only a minority of the cells *were* alive four hours after antibody administration. > A small fraction of the emitters *were* contributing significantly to the overall emission current from the array. ## Discussion section ![image](https://hackmd.io/_uploads/SJzmLsrEJl.png) (PISF: probable in some field) * Move 4 會用比較少 negative language，因為是自己的研究，不是說明自己的 weakness，而是哪些東西不能被 concluded 在自己的研究 ### Phrases of Generality * Overall, * In general, * On the whole, * In the main, * With ... exception(s), * The overall results indicate ... * The results indicate, overall, that ... * In general, the experimental samples resisted ... * With one exception, the experimental samples resisted ... ### Express Limitations * It should be noted that this study has been primarily concerned with ... * This analysis has concentrated on ... * The findings of this study are restricted to ... * This study has addressed only the question of ... * The limitations of this study are clear ... * We would like to point out that we have not ... * However, the findings do not imply ... * The results of this study cannot be taken as evidence for ... * Unfortunately, we are unable to determine from this data ... * The lack of ... means that we cannot be certain .... * Notwithstanding its limitations. this study does suggest ... * Despite its preliminary character. the research reported here would seem to indicate ... * However exploratory, this study may offer some insight into ... # Stylistic Features 1. 通常都用 **We** 而不是 **I** 2. 避免使用 **contraction** * won't -> will not 3. 使用 **formal negative** * not ... any -> no > The analysis didn't yield any new results -> The analysis yielded *no* new results * not ... much -> little > The government didn't allocate much funding for the program -> The government allocated *little* funding for the program * not ... many -> few > This problem doesn't have many sustainable solutions -> This problem has *few* sustainable solutions 4. 不要使用 **vague expressions** * etc., and so forth 5. 不要使用 **you** * you -> can be seen > You can see the results in Table 1 -> The results *can be seen* in Table 1 6. 使用 **indirect questions** 而不是 direct questions > Why has antibiotic resistance increased? -> Many studies have investigated why antibiotic resistance has increased * 可使用 **however**、**nevertheless**、**Despite this** 帶出 question > *However*, this system/process/idea has its problems. > *Nevertheless*, few solutions have been found to ... . > *Despite this*, little progress has been made in ... . 7. 副詞可使用 **mid-position adverb** > This model was developed by the International Monetary Fund (IMF) originally -> This model was *originally* developed by the International Monetary Fund (IMF) 8. **split infinitives (to + V)** > We need to meet the needs of those enrolled in the program adequately. -> We need to *adequately* meet the needs of those enrolled in the program. 9. Using both **active voice** and **passive voice** * 寫 Method 的時候一定是使用 passive voice > Manufacturers inject the virus samples of each selected strain into separate batches of fertilized eggs to amplify the amount of virus. Each virus strain is grown separately inside the eggs over the course of several days, after which it is harvested, inactivated, and purified. # Linking Words and Phrases ![image](https://hackmd.io/_uploads/SyeayY1x1l.png) # Introduction ## Old-to-New ## General to Specific ![image](https://hackmd.io/_uploads/r1wt-KJxJl.png) * Open with Statistics * Open with Definitions * 可使用的 verbs: named, denoted, called, known as, define, referred to, (or ...) * 結構: ![02 General-Specific and Specific-General Text_part01_頁面_14](https://hackmd.io/_uploads/SyCIXYylyg.png) * whereby 後面接一個機制 * **Extended Definition**: begins with a general, one‐sentence definition and then becomes more specific as additional details are provided * Former & Latter 常用句型 * The former deals with ... , while the latter is concerned with .... * Variation in Definitions * While debate exists regarding a precise definition of ... , the stance adopted in this paper is that .... * For the purposes of this paper, ... refers to/is defined as/is considered to be .... * Here we define ... as ... In this paper I have adopted [author's] definition of ... * **This/These** + Noun 做 summary > Writing instructors know that students need to understand the differences between formal and informal language. *This understanding* can help students make strategic choices in their writing. * Definition 內避免使用 **when** 及 **where** > Pollution is when the environment becomes contaminated as a result of human activity. -> Pollution is a form of environmental contamination resulting from human activity. # Problem-Solution 格式: ![image](https://hackmd.io/_uploads/ByH9_F1gyx.png) * Cause‐And‐Effect * **thus + Ving** 可取代 therefore 及 as a result > An increase in demand causes a rise in prices. The tsunami was triggered by a very powerful earthquake. Researchers worldwide are increasingly pressured to publish in English language journals, *thus leading* to a decline in publications written in languages other than English. * Process * Process 步驟之間的連接可用 **participles (分詞)** > First, the virus strains most likely to cause disease are identified and three are *selected* for vaccine development. The virus samples of each *selected* strain are injected into separate batches of fertilized eggs to amplify the amount of virus. * **by + process** 或 **by + non-human-agent** > The impact velocity can be obtained *by calculating* the difference of the arrival times of the two waves. > The increased mobility provided *by this new joint* allows wearers of the finger prosthesis to hold a cup, to pick up a piece of paper, and in some cases to write again. * Solution 常用句型 * Solutions to this problem are now widely discussed. One remedy is to .... * One method to address this difficulty is to .... * There are two possible ways of handling this problem. The first .... The second .... * Several options are available to address this obstacle. However, the best one seems to be ... . * A radically different design/model of ... can overcome this limitation. * Recently, researchers have made significant progress in overcoming this difficulty by .... # Data Commentary Data Commentary structure: 1. location elements and/or summary statements 2. highlighting statements 3. discussions of * Implications * Problems * Exceptions * recommendations * other interesting aspects of the data > Example: ![image](https://hackmd.io/_uploads/ryBcF-LgJe.png) :::info 1. 可用 Google Scholar 配合關鍵字找常用的用法 > "It is * to note that" 2. 在同一個 chapter 內描述 data 及給出見解 ::: :::warning 1. 不要重複圖表上的資訊 > (圖表上寫 76%) -> over three out of four ... 2. 不要講圖表中看不出的資訊 ::: * summary + **location element** with passive voice > The types of internet misbehavior common among university students *are shown in Table 4.* * **Indicative statement**: summarizes what kind of research was done > Table 4 shows the types of internet misbehavior common among university students. **Informative statement**: highlights something interesting about the data > Table 4 shows that illegal downloading of music or films is common among students. :::danger **provide** 只能用在 Indicative statement 而不能用在 Informative statement，且其後 **不能加 that** > :o: Table 5 *provides* demographic information for the study participants. > :x: Table 5 *provides that* most study participants were fairly competent internet users. ::: * **as** ... **in/by/on** > *As* shown *in* Table 1 .... > *As* can be seen *in* Table 6 .... > *As* predicted *by* the model .... > *As* described *on* the previous page .... * Moderating or Qualifying a Claim * 可能性語句 (上 (stronger) <--> 下 (weaker)) * It is clear that ... * It is rather clear that ... * It is very probable/highly likely that ... * It is probable/likely that ... * It is possible that ... * It is unlikely that ... * 可能性形容詞 (上 (stronger) <--> 下 (weaker)) * There is a *definite* possibility that... * There is a *strong* possibility that... * There is a *good* possibility that... * There is a *slight* possibility that... * There is *little* possibility that... * Distance (避免自己的 claim 太強烈) * **seems**、**appear**、**tend** > Children living in poverty *appearseem/tend* to do poorly in school. * **With the exception of**、**Apart from**、**Except for** > *With the exception of/Apart from/Except for* those enrolled in specialized programs, children living in poverty do poorly in school. * weaker verbs > Many studies have concluded that excessive credit growth *caused* the global financial crisis. -> Many studies have concluded that excessive credit growth *contributed* to the global financial crisis. * Based on the limited data available,… * According to this preliminary study,… * Based on previous surveys,… * According to some earlier studies,… * In the view of many scholars,… * ... + according to some recent research * ... + in some cases * Data Comparison * 避免以文字重複資料，而是詮釋資料 (從中找出一些意義) > Almost *twice* as many women as men reported .... > A *smaller* percentage of women reported .... > Nearly *three times* as many men reported .... > Nearly the *same* number of men and women .... > The percentage of men who thought there was only a *slight* risk was over twice that of women > The percentage of women who thought there was *quite* a bit of risk exceeded that of men. > :x: *Fewer* girls than boys reported that their parents decide matters of faith. > :x: *More* boys than girls reported that their parents decide matters of faith. > :x: *Not as many* boys reported that their parents decide matters of faith. * Dealing with Unexpected Outcomes * ... may be due to ... * ... can be attributed ... * ... can probably be accounted for by ... * ... probably a consequence of ... * ... would seem to stem from ... # Summary 別人的 paper 做 citation * 可以用 **paraphrasing** 技巧做 restatement * 開頭句 * The author goes on to say that ... * The article further states that ... * [Authors' surnames here] also state/maintain/argue that .... * [Authors' surnames here] also believe that ... * [Authors' surnames here] conclude that ... * In the second half of the paper, [author's surname here] presents .... * [Authors' surnames here] go on to describe ... * The author further argues that ... * 展示 **相似性** * Similarly, ... * Similar to * Likewise, ... * As in X in Y ... * Like X Y ... * the same ... * as well * beV similar to * resemble * beV comparable to * correspond to * 展示 **對照** * In contrast, ... * Unlike X Y ... * In contrast to ... * On the other hand, ... * ...; however; ... / ... . However, ... * , but ... * Whereas ... , ... / ... , whereas ... * While ... , ... / ... ,while ... * differ from * contrast with * beV different from # Writing Critiques 在對別人的 paper 做完 summary 後，提出 critiques * Critical questions to consider while reading other papers 1. Who is the audience? 2. What is the purpose of the article? 3. What research **questions** or hypotheses are being addressed in the article? Are the questions relevant? 4. What **conclusions** does the author draw from the research? (Hint: Does the author answer yes or no to the research questions?) 5. What kind of **evidence** was collected to explore the research questions? Is there any evidence that could or should have been collected and included but was not? **How good is the evidence?** How well does the evidence **support the conclusions?** 6. Are the author's conclusions valid or plausible based on the evidence? Why or why not? How do the conclusions relate to what you have been learning in class? Are they consistent or inconsistent? 7. Are there any important **assumptions** underlying the article? How do these influence the conclusions? 8. Are the charts, tables, and figures clear? Do they contribute to or detract from the article? 9. Does the research make an original **contribution** to the field? Why or why not? 10. If the evaluation is focused on a published article, **why was the article chosen?** Is the research consistent with material, best practices, and perspectives presented in class? * 提出它哪些地方應該要做但沒有做 * ... could have received ... > This was a well‐written review of current information, but the connection between nutrient stress, secondary compounds, and herbivory rates in wetland plants *could have received* greater coverage. * ... would have been ... had related / had provided > This article *would have been* more persuasive if the author *had related* the findings to previous work on the topic. * if ... had given ... > It would have been better *if* the authors *had given* their main findings in the form of a table. * 使用 past tense verb 表示當前的假設 > Your paper would be stronger if you *included* some additional information. * **should have** 是批評，**could have** 是建議 * success, succeed, successful, successfully / failure, fail, unsuccessful, unsuccessfully > Early papers *unsuccessfully* attempted to use cross‐section distributions of accident counts to distinguish between true and spurious state --- > Early papers *failed* to use cross‐section distributions of accident counts to distinguish between true and spurious state dependence. * tone down: "ambitious, but flawed", Although/However/Despite ... fails/dated(過時) > In this *ambitious, but flawed* study, ... > Despite the many interesting citations in support of his view, the citations are dated and are not likely meaningful today. * ![image](https://hackmd.io/_uploads/Bk8Yhprz1x.png) * 開頭句 * [Authors' names] present a plausible case that ... Less adequate is their discussion of .... * [Authors' names] take on the difficult task of .... Unfortunately, ... * [Authors' names] present an important discussion of .... Although we may not agree on all the issues raised in the article, we praise the authors for .... * While the authors' position that ... is attractive, there are a number of weaknesses in this concept. * The article by [author name] is an ambitious feat of synthesis, encompassing diverse theories of .... This effort, however, is not fully successful. * [Authors' names] have written an important and timely article on .... Despite its many strengths there are a number of small, but important, weaknesses. * [Authors' names] present a compelling argument for ... ; however, ... # Conclusion ## Specific-to-General ![image](https://hackmd.io/_uploads/r1cuDKkx1l.png) # 其他注意事項 * 每個 paragraph 前 **一定要有 opening sentence !!!** * 寫 paper 時要 mimic advisor 的寫作 style * 不要 emphasize 太多 (例如: 大寫開頭、粗體、...)，只標示一定需要注意的地方 * 不要花太多時間在 introduction，而是要想 related work，到底哪些是重點，還有自己的 contribution。 * Related Work 可收集 20 ~ 30 篇以上，自己先篩一下: 先全部的 title 看過，把每篇歸類，再看全部的 abstract 並修正分類，再看 introduction (topic sentences, last sentences)，也可看 conclusion，大概掃過一篇 paper * Background 應和 Related Work 分開 (如果 Background 很多，可分成兩個大 section，否則一般都是分成一個大 section 下的兩個 sub-section)，Background 是做基本知識描述，Related Work 是做墊底，我們的 contribution 需要比他好 * Related Work 只要提及跟自己論文有關的部分就好，不用全部寫上他的 summary * Methodology 與 Technical Approach 差別 * Methodology 類似「使用者觀點」或「使用手冊」 * 不要用 bullet list，也許可以改成用 (i), (ii), ... 寫成一整段 * > It can be observed that the curves show upward trends with some fluctuations on all the datasets. (i) For multi-hop questions in HotpotQA, the standard retrieval is relatively weaker. Complex questions can be not specific search queries and show a larger gap from rewritten queries, i.e., the green and red lines. (ii) On AmbigNQ ... * Figure A LaTex macro: `\FIG`，不要直接寫 `Figure A` * 不要這種句子 (逗點附加，沒有主詞): > As the design progresses, the states of requirements and provisions will dynamically change, *ultimately achieving the fulfillment of the design goals.* * 寫完整的 paper 前，可先寫每個 paragraph 的 **outline**，這樣比較容易知道要寫什麼 * Introduction (引言) * 介紹主題的背景資訊、研究的重要性，以及讀者需要知道的基本資訊 * statement of purpose (目的陳述) * 清楚說明研究或文章的目標和範圍 * justification of the measurement (測量正當性) * 解釋所採用的測量工具或方法的合理性和適用性 * display of data, analysis of the data (資料展示與分析) * 以表格、圖表或文字形式呈現數據 * 解釋數據的模式、趨勢或異常值 * interpretation of data (資料解釋) * 根據數據結果，闡釋其意義，並連結到研究問題 * comparison with prior studies (與先前研究的比較) * 將研究結果與以往相關研究的發現進行比較 * limitations of the study (研究的局限性) * 描述研究中的限制，例如樣本大小不足、偏差、方法限制等 * causal arguments (因果推論) * 討論數據或研究結果是否能夠支持某種因果關係 * speculations (推測) * 基於結果進行合理推測，但可能缺乏直接證據支持 * deductive arguments (演繹推理) * 以邏輯推理的方式，從一般性原則推導出具體結論 * conclusive remarks (結論性評論) * 總結研究的主要發現、意涵，並對未來研究提出建議 # 文法 * cannot 跟 can not 不一樣 * functionality 即表示複數，不是用 functionalities # 口試 ## 投影片 * 看一下你的結構: 每一張投影片的標題不要重覆。要精確的讓聽眾知道現在講的是什麼。不要十張heading都是 Motivation，明明講的是不太一樣的，但是heading沒有辦法分辨。不要expect聽眾會看你全部。有時候看到heading，其他用聽的，或者其他看你的圖. 你至少要 Motivation (1/2), (2/2) ..或是 Motivation: 再加一個片語，引導聽眾 * 內容bulleted list的第一層 不應該只有單一個bullet。如果只有一個，那就應該把它 promote 到heading. 但是 heading 也不可以太長，愈精簡愈好 * 更不OK的是只有單一個 top-level bullet, 單一個 2nd-level bullet, 甚至單一個 3rd level bullet。重構一下，讓它不會那麼不不均、不對稱 * 每個bullet 原則上不要超過一列 (one line)，更不要用整句。聽眾不可能讀你整句的 bullet，因為還要聽你講。應該要切成bullets 和 subbulets，用片語 * bullets 一般兩層最好，三層就極限了，非必要就不要用到三層。不要用到四層。 * 儘可能用圖來 illustrate 概念。圖裡面的字體不要太小，也不要太多東西。如果複雜的結構，要抽象化，附上 label, 必要的時候展開來，再highlight一下你要聽眾聚焦的部位 * 每張 內容slide (不包括outline) 至少講個半分鐘，不然聽眾還沒來得及看完和聽完就過去了 * 上過我的科技英文的都知道，不要隨便用誇大的副詞像「非常」，「超級無敵」。中文的「很」就已經是 "very"了，而且英文也不要隨便用 very。但是中文因為少了 linking verb, 「很」就被濫用在 linking verb 的角色，失去了 "very"的強度。所以現在因為呷重鹹習慣了，常常就是以「非常」來取代 very，甚至「非常非常」-- 我是不接受的！「非常」是 unusual或unusually ，非常非常是 unusually unusual，表現出的就是缺少詞彙、缺乏表達能力。我最不能接受的是「非常正常」-- 自己想想看自己在講什麼！unusually usual!?? 其實不止在科技英文，我在其他文件也都講過不止一次。 * 不要假設聽眾都知道你講的技術面的東西，尤其跟嵌入式系統有關的。你要回想一下兩年前的自己能夠聽懂你自己的talk嗎?? * contributions 不是「你做了什麼 (technical approach)」而是比較 high level「你對領域帶來什麼(1) 新的 (2) 有價值的 (3) 解法不明顯 的解決方案」。要能夠很清楚的在一頁投影片講完。如果講了三、四頁很有可能就是在講 technical approach 或 implementation。 * 不要隨便用「證明」(prove) 和「證實?」(verify) 因為在 CS 是有特殊意義的。prove 是要能夠以數學或邏輯來證明。如果你只秀出了一個案例，那也只是一個 existence proof 但是沒有證明所有的case 都可以適用。如果你做模擬，業界稱之 verification, 但是學界稱之 validation (驗證?), 就是你測資測出來是可以的，但是測資夠不夠完整，並不知道，也不能直接宣稱你已經 verify了。verification 指的是正式的以數學或邏輯證明的，所以是很嚴格的。