I have got a somewhat general question about scientific writing to discuss.
I think, everybody agrees that the paper is the major output of a scientist's work, so producing perfect papers is the primary goal of a research fellow. But what is a perfect paper, and how can we judge about the quality?
There are various criteria, like number of citations, number of downloads, etc, but they all refer to already published and discussed papers. Imagine the paper is newly born. How can a scientist write a nice paper? How it is recognized by the editor? How it is graded by the scientific community?
Thinking of this, I have come to a naive conclusion that a "perfect" paper means different things for different people.
First, the author as a paper producer. He/she wants to present own striking results in an interesting, attractive, and clear manner. From this point of view, writing a perfect paper is a matter of collecting proper data, analyzing it, and presenting it in the logical, properly illustrated and conclusive manner. This is more or less clear, and has been discussed in a number of lectures, books, and publisher workshops.
An author as a member of research department wants to advertise the institution and to get a huge number of citations: good or bad, but now the quality of a researcher and thus the future funding are getting more and more dependent on the formal citation criteria.
Then, an editor wants basically the only thing - to maximize the impact factor of the journal while keeping the scope within preset frames - this means merely less papers and more citations needed. An additional attitude might be to minimize the paper length at the same time, however, with Supplementary material available online this gets minor issue, I think.
As most of the scientists may play different roles (author of the own papers, reviewer of the others' papers, research fellow using the results of others' papers as inspiration, or even editor accepting or declining papers), at different moments he has to analyze the papers quality from different points of view.
So, my questions are as follows:
- When you are writing your own paper, you are the best expert in your particular field, and likely you know well some broader context. How do you make your research appealing and exciting to others? How to provoke the widest possible discussion of your work? How to make the paper interesting even to people outside you particular area (which may lead to fruitful multidisciplinary work)?
- Staying in the frame of the certain scope of the paper (it is more or less fixed by your work already done), how can you inspire more citations of your paper? Imagine that your work is not worth publishing in the TOP journal at the moment, and it will appear as one of the tens or hundreds of papers published this month in your field. How can you attract more attention from your colleagues?
- Imagine that you are a journal editor (again, not of Cell of Nature, but of a middle-class journal), and you receive a paper submitted. The paper is not directly related to your own research field, and the broad context may be unclear. How can you estimate whether the paper will be hot and well cited, or it will be hardly noticed in spite of the properly done research? (Assume for a moment that you don't have ideal peers, or their opinion is unclear, and you should decide about the acceptance yourself).
I would be glad to hear your opinions (especially the point of view of practicing editors is interesting for me). If the topic has already been discussed elsewhere, could you please give me a reference?
I have got some ideas on the subject, but I would prefer to share them after I get some feedback from others.
1. Are the results new?
2. Are they important?
3. Does it overturn our previous thinking?
4. Is the research trustworthy?
5. Is it reported clearly?
Citations and impact factors are poor proxy measures of the above.
To make the paper appealing, use photographs, graphs, equations, text and tables to get the point across as painlessly as is possible.
Obviously hold as many seminars as you can on your work, whenever you can.
Do not be afraid of "marketing" your paper through popular magazines and newspapers, but this will not gather citations.
I am not in favour of "trying to get people to cite your work". You can take a horse to water, but you cannot force it to drink. Asking people to cite you is unprofessional.
Nevertheless, the greater the exposure your paper has, the more likely you are to garner a citation. So publish it on the Web. Give out free physical reprints at every occasion. {Pretend that you only have a few copies left, and you are doing the person a big personal favour in giving them your 'last copy'!}
To gain the attention of the journal editor your research should be revolutionary rather than evolutionary. This goes back to my question 3 above.
Yes, there is a "game" being played, but the ethical rules are strict.
1. Are the results new?
2. Are they important?
3. Does it overturn our previous thinking?
4. Is the research trustworthy?
5. Is it reported clearly?
Citations and impact factors are poor proxy measures of the above.
To make the paper appealing, use photographs, graphs, equations, text and tables to get the point across as painlessly as is possible.
Obviously hold as many seminars as you can on your work, whenever you can.
Do not be afraid of "marketing" your paper through popular magazines and newspapers, but this will not gather citations.
I am not in favour of "trying to get people to cite your work". You can take a horse to water, but you cannot force it to drink. Asking people to cite you is unprofessional.
Nevertheless, the greater the exposure your paper has, the more likely you are to garner a citation. So publish it on the Web. Give out free physical reprints at every occasion. {Pretend that you only have a few copies left, and you are doing the person a big personal favour in giving them your 'last copy'!}
To gain the attention of the journal editor your research should be revolutionary rather than evolutionary. This goes back to my question 3 above.
Yes, there is a "game" being played, but the ethical rules are strict.
1. Does the research have any value?
2. Is it interesting to most people or to a particular group of people?
3. Does the research methods or experimental procedure trustable?
4. Does it have a concise, accurate, and attractive title?
5. Does the author present it in an easy understanding and logical way?
6. Does the author illustrate the major contents in vivid and beautiful figures (pictures, graphs, and diagrams) or clear tables?
There are many ways to increase the value, readability and its impact of a scientific paper .
Thank you for responses, They are very well in line with my understanding: clear idea, logical writing, explaining illustrations, and strong conclusion make the paper better, even if one is limited in facilities and cannot improve the contents of the work.
@ Ian Kennedy - of course, you are right about ethics of the forced citations. I did not mean that, my concern was how to make the paper appealing to cite it - that is, not force, but persuade.
Let me go into detail of the other part of my question. Besides my own research, I do some editing in a journal. It is not extremely popular, the scope is very general, and (most important) the papers are normally submitted in Russian. (They are later translated intio English for publication, by when deciding on acceptance, the editor holds Russian text).
Thus, 1) the submitted papers are very diverse, covering all the chemistry-related areas, and 2) the pool of peers is not too huge. Often ALL Russian-speaking specualists in the particular area belong to the submitting research group. This of course complicates the review process.
Imagine:
You deal with inorganic synthesis of special copper-based complexes with interesting catalytic and photochemical process. For review, you get a paper on similar complexes, but silver-based, and their application in analytical determination of certain pollutants is commented, along with some antimicrobial properties. You see then that the paper is writeen quite clearly, and there are now significant mistakes in data interpretation - so imagine that you trust the scientific scope of the paper.
Even though the paper topic is in the field of your research, you cannot say at once, whether the result is 100% original, partially new, or just duplication. Is it at all important for determination of those pollutants? Is the observed antimicrobial effect significant for medicine, or there are cheaper analogs? I assume that to answer these questions you will need to dig deep into the topic, and the time is what we always lack...
So my question is: when you read a paper related to your branch but not 100% in your area - what "signals" do you look for to understand how good or bad the paper is? For example, I consider the papers with >50% of self-citations poorly supported (if this is not a review, sure). Also, if I cannot locate the aims and goals clearly stated BEFORE discussion, I consider the work "experiment to just do som experiment" - and I do not recommend the paper for publishing as is. Is the spectral data is dublicated in the text, table, and in the Experimental for no special reason - this is also a signal to me that the paper was written without much efforts on the text.
Can you share the signals that you use (maybe, unconsciously) to judge about the quality of the paper you read?
If care and attention has been taken in providing meticulous referencing, I am initially inclined to give a favourable review, as care has been taken with the final details.
Evgeny - you give this topic a lot of thought. It sounds to me that , when editing, you have your 'cautious radar' on. As an editor and reviewer for over 30 journals - I think that is a very important quality. I suppose that good editors often have a wealth of experience - or learn the job quickly. I think that there is a lot to be said for having a 'gut instinct' for anything that raises suspicion. OK - a few will 'get through the net' - but, on the whole, most don't get through to publication. Poorly structured papers are often the first warning sign.
@Dean Whitehead
Thank you for your comment.
You are right - I am quite new to the scientific editing field. Most of the papers submitted to our journal do not follow the quidelines in this or that aspect (format, illustrations, references, poor introduction, or whatever). However, it is impossible to reject all such papers - or we will have nothing to publish. Thus, the accepted practice now is such that if there are no extremely awful mistakes in the text, the editor does the best to reshape and correct the paper in scope of the reported data. This takes mich time and increases the production time of all the papers sufficiently. This is why I am trying to develop the criteria (for myself) to "worst of the acceptable" papers for editorial and return "best of the poor" ones to the authors for revision. So, whereas the definitely good and deadly bad papers are quite easily distinguished, but in the middle segment the decision is complicated...
Again, Evgeny - I repeat that you are obviously taking your role seriously and trying to do the best for the journal that you represent. True, the 'middle segment' will always be the most difficult (and, of course, the most in number) - but I would argue that, because you have more easily separated the good from the bad, most will be 'reasonably acceptable'. The trick, in this situation, is to be able to grade or prioritize them - usually on the basis of originality and/or topicality.
A quick trick to check the structure is to open the paper with Open Office Writer, and to generate a table of contents automatically. This quickly reveals any problems in the structure of the paper. At the same time, use the spelling checker and a grammar checker to see if those aspects have already been attended to.
Glenn, call it ego-centrical, but all my work is genius until someone convinces me otherwise - and they have to work hard to do that ; ). On the other hand, my family keep me grounded by reminding me that 'it's never going to set the world on fire' - and that Dr Who, Top Gear, Sherlock etc are far more important things in life.
Yes - keep it real and straight-forward. No smokescreens, no hiding behind pedantic and/technical jargon, no 'writing to impress'.
I think, there is no perfect research paper, but it is a good paper with standard requirements. If you mean a perfect research paper which fits with the scop and instructions to author, it is ok. Otherwise, papers which published in science or nature are considered perfect others are not. Again good paper but not perfect.
Correct Glenn - highly cited works are always a good indication of content - but probably also related to good presentation as well. The two, for me, usually go hand-in-hand. I would argue that it is too problematical to list a 'top ten' great papers because there are so many different disciplines, too many methodologies - and some perceptions of 'great' may well be personal and subjective. In my field, I recently reviewed an article that was about the top cited articles. The main distinguishable feature was that most of them were seminal papers - often written 20-30 years ago. To present them today as a template may be problematical as well. Writing styles and journal house-styles change over time.
In addition, the attached article entitled ''What makes a good research paper?'' by Hughes W. may be useful.
Regards
http://centaur.reading.ac.uk/4306/1/ARCOM_Newsletter_-_What_makes_a_good_research_paper.pdf
Perfect research paper must have applications. The steps to write axperfect paper are presentedd in the file
- Badges
- Science topic
- Similar topics
- Education
- Higher Education
- Academic Writing