I think that the number of citations does not reflect the essence of novelty publication. Can be noted 100 papers without saying anything, but you can 5-6 and note the most important aspects of the study.

I think that the number of citations does not reflect the essence of novelty publication. Can be noted 100 papers without saying anything, but you can 5-6 and note the most important aspects of the study.

I read somewhere that influential papers at cited at least 50 times in a 10-years period.

These 5-6 are major publications . So you need to choose the main publication for citations. Yes it is hard to do. But it shows your knowledge of the problem.

In this case, the amount does not determine the quality.

Dear Vyacheslav,

So you just take the 5-6 most cited publications for a given author X to have an idea about the impact of author X, independent how many author X published in total, right?

If it involves the consideration of one problem, yes.

Interesting question. It depends on the subject of the paper. If you are citing a newly born idea, few citations would be good enough. But to quantify exactly a number seems difficult. I have seen papers where more than 50 publications have been cited but it does not mean it would make a publication credible. I agree with @Vyacheslav Lyashenko, that "number of citations does not reflect the essence of novelty publication". The essence is the critical discussion on the cited papers.

What is the reason for this deterministic breakthrough, dear Marcel?

—g

Dear Giuseppe,

what do you mean exactly in simple words?

One could argue that when a problem is solved, research on that problem stops, so citations dealing with the problem also stop. The more citations concerning a publication dealing with problem X, the higher the chance that problem X has not been solved?

Dear Marcel. But you can also say that the problem is interesting and has a lot of decisions. Then the task is to find the optimal solution, and not the overall solution.

Even if a problem is solved, the approach applied to address that specific problem can be re-used in solving another problem(s) in other parts of the world.

First of all, these are mathematical, statistical analysis techniques are used to solve problems in different areas. As an example may be called stochastic frontier analysis. It is used for the analysis of various economic data, sociology, situational analysis ...

Dear Marcel,

Number of citations required to call a cited publication influential depends on the research and the domain of the research itself.

@Marcel M. Lambrechts,

Implementation of Spatial Data Infrastructures, E-government , precision farming and Disaster Management etc are some of the examples that can be cited immediately.

What about fundamental research problems? In evolutionary biology it is accepted that natural selection occurs or that species result from evolutionary processes including natural selection. How often is the first publication of Darwin cited or ignored in current scientific publications dealing with evolution and natural history studies? For some, natural selection is so obvious that citing Darwin in a current publication dealing Evolution, Evolutionary Biology, Evolutionary Ecology, etc.might be considered as old-fashioned..... How often Mendel is cited in current studies of Genetics or Quantitative Genetics....

Dear Marcel,

I guess 800 or 850 citations per author per article. I read it somewhere before, i dont know the reference now.

People can not only be cited in a written publication, but also during an oral presentation or a discussion? From what fraction of the objects we have at home can we cite the inventor?

Dont panic we are both long away from those numbers.

If you are speaking about MY paper, then "one" citation is sufficient (for me) to be considered 'influential". If you are speaking about my colleague or friend, then "two" would be necessary. For a complete stranger, I would require "three" citations. And if you are working in a competitor's lab, you can never get enough citations! (Of course I speak only jokingly). A genuinely influential paper will change the course of research and application in that field (whether pure science, medicine, sociology, etc) for at least a decade if not longer. So it must take at least 10 years or more to determine if a published article is really considered "influential".

Dear Marcel,

The same thing as you said "being frequently cited can be either a good or a bad sign. It depends on the content of papers and why the papers are cited...." can be concluded from up-voters in RG, too!

No, it is not determinable, Marcel.

Same as for that old question here on RG, about the lifespan for a publication or a discovery, its expiration date. Who knows? Do we know how many implications could a new statement could propagate, into an open unlimited system?

We could set some indicator value — than use the consequent standardized scale.

But in the real things it is not predictable, and it should not be allowed — or, publications may propagate in derived strings, and be considered "goods" to bet on or to be used as "currency". At that point, the whole system of research would become corrupt, because a large number of sub-systems should seek to win myriads of partial bets — "fractions”.

BTW, a certain study I've read some time ago calculates that the most influent authors in the history of medicine of the last 50 years, have >500 papers published. Much less for sciences not related to industries [great philosophers of science with

But then again, we talk about quantity, not quality of citations. In another related question, people apparently prefer 'quality' more than 'quantity' of publications. Perhaps the same could be argued for citations and the reasons for citations? Quality of citations might be more important than quantity of citations.

Fictive example without judging people:

Better 5 citations from a famous experienced researcher than 200 citations from inexperienced researchers, or not? Better 5 citations that support your idea than 200 citations that criticizes your idea? Etc...

@Giuseppe Laquidara: No, it is not determinable…

Agreed!

As an indicator of the influence of a publication, citation rate will vary, depending on the discipline and the obscurity of a journal topic. For example, a paper on convex hulls or relators would be of interest to a very small community.

An overview of rankings of universities and scholarly output is given in

A. Marcus, On the mathematics of ranking universities and scientific products:

http://conference.ubbcluj.ro/competitiveness/files/marcus_math-of-ranking.pdf

Dear @Marcel, this is a good question and a controversial issue. I support the answer of dear @Vyacheslav that the number of citations does not reflect the essence of novelty publication. In many cases, there are very strong and novel methods but the people do not go back and cite them because they are heavy, deep, and dense, and, therefore, difficult for many readers.

I can't recall a reference but I have seen somewhere a statement that a Nobel-prize-level paper in physics should gather ca. 1000 citations.

Dear Marcel. Quality is very important quotes. But sharing quotes from pilot studies degree probably not worth it. In this case, the work itself is important, not its author.

As for citations supporting your work. One can construct an overview that any quotations will support your work. A contradiction only will emphasize the importance of research.

And yet, quantification of citations to achieve that either (getting recognition, awards) are only worsens the research itself.

Dear Vyacheslav,

A very interesting remark. We often cite articles referring to (an) author(s) without specifying why the articles were cited. Citing sentences from other articles would provide much more precision why the articles were cited and would also indicate the cited paper was read in detail (e.g. author X wrote sentence X on page X).

There is no threshold for fame. Number of citations does not directly imply influence.

Dear Vyacheslav,

easy question, but no easy answer. Probably, it will not be possible to prove a direct correlation between scientific "Impact" and the number of citations. Very often, the highest citation numbers are found for technical reports. This is reasonable, and might indeed be related to scientific merits, if such a paper describes a fundamental technique for the first time. However, very often, modifications of pre-existing techniques are those that are highly cited, just because they give a reasonable description of procedures and recipes. So, they are cited more due to convenience than to scientific originality and merits.

Many very fundamental papers are rarely cited. However, they have initiated subsequent studies considerably. After a while, especially basic findings seem to be rarely cited, just because the results from such papers are common knowledge, so self-evident and taken for granted that the underlying papers go without saying.

My recommendation: Let's not look too much on measuring scientific contributions by secondary criteria, citations and (even worse) by the amount of money that has been obtained for research, but by reading the publications and by thinking.

Interesting debate - jw

Dear Johannes. Totally agree with you.

The importance of a research article is routinely measured by counting how many times it has been cited. Hence, number of citations appears to be directly related to the paper's quality of content. However, the number of citations received by a paper depends more on the aspect and time when the paper is published. Papers published early in a new field receive many more citations regardless of content (so-called "first mover advantage") than those published later on. Earliest publications in a comparatively new field normally get a head start that is subsequently amplified by the preferential attachment process and they will continue to receive citations indefinitely at a higher rate than later papers. Also papers appeared in high IF journals also receive high citation though IF varies considerably across disciplines (Math has an average of 0.9 citation per article; Life Sciences have an average of 6.2). Using the impact factor alone to judge a journal or articles published in it is like using weight alone to judge a person's health. There are other factors and considering all those, it is quite tricky to arrive at conclusive number of citation to label a cited publication influential.

This is an extract from The Scientist, Dec. 2012

"In the top ten scandals for science this year, an astroturfing organized by prof. (omissis) fabricating >10,000 citations for his new revolutionary theory."

1st comment on the footer: "Fabricating? Give citation demonstrating fabrication…"

2nd comment: "Who says the scandal deserves to be in the top 10?"

3rd comment: "Pls, could the author of the article demonstrate effectively the "omissis"…?"

—g

@ YOGESH CHANDRA TRIPATHI, well said.

Interesting question to ask. However going through the answers, especially the one by @Artur mentioning 1000 citation for a physics nobel prize winning paper, I would ask if these citations happen before or after the Nobel prize. What I mean is now a days it is a bit common to see citations to an individual rather than to a work. Famous people get cited more often than the papers. However, a good work is supposed to draw attention and citations. I would think google scholar refering an index i10 which means 10 citations are also considered good enough.

I agree with @Vijay that  it is a bit common to see citations to an individual rather than to a work.

I think it has to do with the way it is cited, more then the number of citations, some papers propose a novel way of seeing a phenomena or revolutionary way of solving a problem. you can see that it triggered a trend in the field, and how do you do you know that ? By looking at the relationship between the cited paper and where it is has been cited ! if you see that the cited paper was exploited somehow to create something new in the field ! that meas the paper has an influence on the field !

Number isn't always the key element to judge the impact of a paper, Cause usually review papers are one of the most cited papers in every field, but only because they are a good entry to the a field and they are good to discuss the novelty of the proposed work, and perhaps cause they need a lot of efforts and research which is something alot of researchers avoid, making Review papers kinda rare ! So they get cited alot, but this doesn't mean they have a big influence or impact of the field.

That's just my opinion !

NB : Sorry for my English.