Roadmap recall Previous courses (General Writing Skills) Course-1: - - PowerPoint PPT Presentation

1

Roadmap recall

• Previous courses (General Writing Skills)

– Course-1: Preparation, Writing Tools – Course-2: General Issues before writing – Course-3: Word Order, & Readability – Course-4: How to structure paragraphs – Course-5: Break up long sentences (1st+2nd) – Course-6: Being Concise, & removing Redundancy – Course-7: Avoid Ambiguity – Course-8: Clarify Who did What – Course-9: Highlight Your Findings

2

• Today’s class: Course-10~11

– Course-10: Discuss your limitation – Course-11: Hedging & Criticizing – Course-12: Plagiarism & Paraphrasing

3

Course -10:

Discuss your Limitation

主讲：黄华威 副教授 数据科学与计算机学院 学院个人主页： http://sdcs.sysu.edu.cn/content/4989

《科技论文的阅读与写作》课程主页：http://inpluslab.com/paperwriting

4

• Once we accept our limits, we go beyond them.
• - Albert Einstein

5

10.1 Recognize the importance of ‘bad data’

• To enable referees to judge whether you have made a

mistake or not, you should – NOT hide any negative results – be clear and honest about the limitations of your methods and approach

• In Why People Believe Weird Things, author Prof. Michael

Shermer writes: Usually negative findings are – not wanted, and not published, – (but) helping us get close to truth

6

10.2 Be constructive in how you present your limitations

• Try to be constructive, so others can learn more
• Your results may be negative to you, but not for

scientific community

• However, don’t be too negative when presenting
• S1. [Original]

§ The limitation of this paper is that the two surveys were unfortunately not conducted in the same period. This will affect

• ur results in terms of …

Too negative !

• S2. [Revised]

§ Although the two surveys were not conducted in the same period, this will only affect our results in terms of …

7

10.3 Be constructive in how you present your limitations (cont.)

• Avoid to use

– Adverbs: Regrettably, unfortunately – Link words: Moreover

• Change negative link words e.g., moreover, to

– in addition, further, furthermore, also, etc

• S1. [Negative]

§ You are the worst student I have ever had. Moreover, you appear to know absolutely nothing.

• S2. [Positive]

§ You are the best student I have ever had. In addition, you appear to know absolutely everything.

8

10.4 Clarify exactly what your limitations are

• Example
• S1. [Fail to be clear] One limitation of our research was the sample size,

which was too small.

• S2. [Fail to be clear] The unfortunate contamination of a few of our

samples may mean that some of our conclusions are somewhat misleading.

• S3. [Clear] One limitation of our research was the sample size. Clearly

200 Xs are not enough to make generalizations about Y. However, from the results of those limited number of Xs, a clear pattern emerged which…

• S4. [Clear] Two of our samples were contaminated. This occurred

because … We thus plan to repeat our experiments in future work. However, our analysis of the uncontaminated samples (24 in total) supported our initial hypothesis that .

Discuss possible remedies (补救) Honest Clear

9

10.5 Refer to other authors who had similar problems

• A way to lessen the impact of your ‘bad data’ is

– to say that other authors have experienced similar problems

• However, your story should be convincing

– Explain the pitfall (缺陷) – Give reason for the pitfall – Outline consequence of the pitfall – Refer to a similar pitfall experienced by another author [Example] § The statistical tool may not be optimal for describing some of the variables involved. However it is optimal for x, y and z. In addition, exactly the same tool was used for conducting similar research with an American sample [Williams, 2017]. Williams’ results were reliable and representative and were in fact used by the US government.

10

10.6 Explain why you did not study certain data

• Cases include

– You did not study the most recent data – You did not study a sufficient amount of data

• How to deal with them?

[Example: Old data was used] § Even though the data were collected two years ago, the stability of this sector means that such data have not changed significantly. In fact, in the last two years the percentage of x has remained exactly the same [Wang 2017, Chu Wa 2018]. In addition, more recent data are not currently available. [Example: Insufficient data was used] § Our data only refer to one kind of sector. However, as far as we know there are no similar studies for this sector in China. Thus we believe that this project opens the way for …

11

10.7 Tell the reader from what standpoint you wish them to view your data

• To clearly express your point of view, you may use

– in our view and we believe

• Tell the reader from which standpoint you want them to

interpret or judge your data

[Examples:] § Viewed / Seen in this way, the data take on a different meaning. § From this alternative perspective, these findings shed new light on … § From an X point of view, the results can be interpreted very differently § In this view, these data may mean that … § Under these conditions, it is reasonable to pose a new perspective on … This technique helps distance you from your own data.

12

10.7 Tell the reader from what standpoint you wish them to view your data (cont.)

• Not only data, but also your method, model, discussion,

hypothesis etc.

[Examples:] § These data indicate that … § The evidence favors the conclusion that … § The model predicted that … § From this discussion, it would appear that … § The hypothesis seems plausible because… § The existence of such phenomena may give confirmation of … [Useful verbs in such contexts are:] § Imply, § indicate, § suggest, § point toward, § hint at § …

13

10.8 Don’t end your paper by talking about your limitations

• Don’t end your Discussion / Conclusions by

– talking about your limitations

• End with something positive

– this will be the reader’s final impression

[Possible endings are:] § talking about other applications for your findings § suggesting other avenues of research for the future § re-highlighting the benefits of your findings This technique helps increase the acceptance-ratio of your paper.

14

Course -11:

Hedging and Criticizing 防备 与 批评

主讲：黄华威 副教授 数据科学与计算机学院 学院个人主页： http://sdcs.sysu.edu.cn/content/4989

《科技论文的阅读与写作》课程主页：http://inpluslab.com/paperwriting

15

Why is this chapter important?

• This chapter helps you

– (1) learn to predict possible objections to your claims

u This is to make claims about your findings in a way that the referee

and the research community are more likely accept them

– (2) criticize the work of other authors in a constructive manner by

u building upon their findings,

u rather than underlining their inadequacy

Hedging (防备) Criticizing (批评)

16

Ø Hedging (防备)

• A hedge was originally a fence or boundary delimiting an area of land (划清

界限)

• Here it means you protect yourself against some adverse risk
• Hedging requires anticipating possible opposition from your referees and

readers, by not saying things too assertively or directly.

Ø Your risk while writing too assertively

Ø criticism by referees and other researchers

Ø The ideas:

1) Express yourself with honesty, precision and caution

2) Become diplomatic (老练的) in any criticisms

11.1 Why and when to hedge?

17

Ø Hedging (防备)

• does not mean that you should be vague (含混不清的)
• Instead, as precise (明确的) as possible

11.1 Why and when to hedge? (cont.)

[OV:] § Although many authors have investigated x problem in previous studies, this is the first attempt to systematically analyze x+ problem . [RV:] § Although many authors have investigated x problem in previous studies, we believe / as far as we know / to the best of our knowledge this is the first attempt to systematically analyze x+ problem . [OV:] § Our results demonstrate that … [RV:] § Our results would seem to demonstrate that …

18

11.2 Highlighting and Hedging

• Highlighting (高亮、凸显) and Hedging (防备)

– are not contradictory skills – In fact they should be used hand-by-hand [According to previous course, Highlighting means:]

§ Helping the reader to see your findings easily (not hiding them in the middle of a long paragraph) § Using shorter sentences when giving important information § Using more dynamic language when drawing attention

* You can do all the above and still hedge.

19

11.2 Highlighting and Hedging (cont.)

• Compare the following 3 sentences
• S1. This is a very important finding.
• S2. These results suggest that this is a very important finding.
• S3. Interestingly, these results prove/suggest that X is

fundamental in producing Y. Here the skill is in

• finding the right balance of highlighting and hedging, and
• knowing how to hedge so that referees think you are sincere

20

11.3 Toning down verbs

• Toning down (缓解): Leave room for doubt

[Verbs that are too tough:] § This factor is responsible for the increase in… § These results demonstrate the importance of … § These findings proves that x=y. § This problem manifests (表明) itself in … § This means that x=y. [RV:] § This factor is probably / may be responsible for the increase in… § These results would seem to show / indicate / suggest the importance of … § These findings provide some evidence to prove that x=y. § This problem tends / seems / appears to manifest itself in … § This means It seems likely / probable / possible that x=y.

21

11.4 Toning down adj and adv

• Some Adjs / Advs have a very strong tone (语气、色调)
• Carefully use these Adjs and Advs, when

referring to your own work; otherwise, risks arise

[Innovation:] § Innovative, novel, cutting edge (尖端的，最前沿的), seminal (有 创造力的，对未来有影响的；重大的), pivotal (关键的；中枢 的；枢轴的)… [Importance:] § Extremely important, very significant, of fundamental importance… [Certainty:] § Clear(ly), obvious(ly), evident(ly), conclusive(ly), definite(ly), undeniable, undeniably, undoubtedly…

• S1. * This pivotal approach is particularly interesting for data scientists.

Arrogant (自大、傲慢) Do the data scientists think so?

22

11.4 Toning down adj and adv (cont.)

• Risk: carefully use these Adjs and Advs, when

referring to your own work (cont.)

• S2. Our approach would lend itself well for use by data scientists.

lend itself to ( )

• S3. We hope that data scientists will find our approach useful.

23

11.4 Toning down adj and adv (cont.)

• Certainty, useful advs and adverbial phrases

– Somewhat, to a certain extent

• S4. X is related to Y.
• S5. X is certainly related to Y.
• S6. X is somehow related to Y.
• S7. X is likely related to Y.

More indirect approach

24

11.5 Inserting adv to tone down strong claims

• Different Advs indicate different levels of

confidence

• S1. X is clearly visible.
• S2. X is scarcely detectable.
• S3. X is reasonably clearly visible.
• S4. X is scarcely detectable, at least in our experiment.
• S5. Our data fit perfectly with those of poisson distribution.
• S6. Our data fit quite well with those of poisson distribution.

§ Other words to replace quite: surprisingly, remarkably, and unexpectedly § Use significantly wisely: it is often associated with sth is unlikely to have occurred by chance

25

11.6 Saving your own face: revealing and obscuring your identity as the author in humanist subjects

• argue (争论，辩论；提出理由)

– Authorial voice: state sth that may go against what other authors have previously claimed

• S1. I argue that ...
• S2. The present study argues that ...

suggest, infer, calculate, and believe

26

10.7 Saving other authors’ face: put their research in a positive light

• It is fine to question other’s findings and

conclusions

• But when making criticisms, be sure to put the
• riginal research in a positive light
• In this way, you save the original author’s face,

– i.e. their reputation and position in the academic world

• Suggestions:

– Consider when the referee & reader of your paper are reading your criticisms – In a constructive way

27

11.7 Saving other authors’ face: put their research in a positive light (cont.)`````

• Structure your criticisms
• S1. A lot of link failure-recovery solutions have been presented

in the literature (refs). Those existing studies indicate that it is of great importance to protect the controller-switch channels in software-defined networking.

• S2. However, Most of the previous studies focused on the

handling of reactive single-link failures. Thus, the proposed algorithms can only address the single-link failures in a reactive way.

• S3. However, multiple concurrent failures are pretty often in

nowadays networks. This is why in our study the non single-link failures in a proactive way.

28

11.8 Don’t overhedge

• Be careful

– not to follow a strong positive assertion with a weak statement that undermines it (S1) – avoid having several levels of hedging (S2)

• S1. It is clearly that yellow may be preferable to red for alerting

danger (警示危险).

• S2. It may thus, given these particular circumstances, be assumed

that there is a certain possibility that yellow may be preferable to red for alerting danger.

• S3. It is clear that yellow is be preferable to red for alerting danger.
• S4. It may thus, given these particular circumstances, be assumed

that there is a certain possibility that yellow may be preferable to red for alerting danger.

29

Advance notice

• Thank you so much for your attending!
• Next class:

– Course-12: Plagiarism & Paraphrasing – Course-13: Sections of paper

u Title of your paper