Kamal H Karim started the discussion here in RG six month ago: “Why there are too many false citations in research papers published in international and local journals?” My impression is this process is boosted exponentially during the last decades. Many causes reasons and problems have been discussed here by many researchers. Many participants in discussion said: Responsible for a correct citation are primarily the authors, but also the peer-reviewers and the publishers have a share of responsibility. Many participants in discussion united the desire that it should become better again. The implementation, however is difficult for many reasons like: 1. A culture change in universities of the past 2 decades has been; students are expected to start doing research in their first year. In some cases, they are required to have publications before graduating. This puts more pressure on students to do things faster. They feel they don't have the time to spend doing things the 'long' way. 2. In the age of Big Literature we can see knowledge is published in prolific rates. The magnitude and rapid expansion of literature makes it increasingly impossible to conduct comprehensive and transparent assessments. During my study at the university and my first scientific work in the 1980s I have learned by my old professor that I have to read all literature sources entire, not only the summary and I have to read always the primary sources of literature, the original paper, not only the cited extracts in a secondary source. I worked so in this way over all the years and I’m doing it in this way today. But this means, I’ not able to write and publish 10 or more papers in the year as some researchers do. However the evaluation of researchers and the allocation of funds for projects often depend on the quantity of published papers. That’s not a good development and it might contribute to false citation and plagiarism. Somehow, there is often a lack of time to be thorough and conscientious and that’s not good at all. Can modern technology help us with this? Can computer algorithms now make fewer mistakes then we do or make more mistakes? On the one hand at the moment it is often still this: When we use automatic literature search systems like Web of science advanced search or Scopus, we get many hits, but many of them turn out to be useless because they do not really contain what was queried. It is still difficult to rely on it. On the other hand: Innovations in research and assessment practices and tools are needed. We need gig data methods for analyzing literature and making assess urgently. But caution: we also should ensure comprehensiveness. This is a problem really. What do you all mean: In the age of big literature, can new computer-aided methods and procedures reduce or increase the numbers of false citations in research papers?
Dear Petra,
I fully share your concern about the trade-off between "quantity" and "quality".
I believe that a high quality work can not be compatible with the average production required of a researcher today, at least if we talk about the highest possible quality.
You're right in suggest that the role of Big Data and Data Analytics, in this context will be increasingly crucial.
About this, at a general level, I would suggest a book (Cathy O'Neil, Weapons of Math Destruction. How Big Data Increases Inequality and Threatens Democracy):
although the book doesn't speak directly of your topic, it is quite eloquent in show the basical limits of such models, that are applied also in this case.
Best regards,
Marco
Thank you Marco, for your inspiring answer! I think I will read the book because these questions generally move me: Possibilities and limitations of new technologies. How much control must and can humans keep about processes? Can models, computers and algorithms really everything as more and more is suggested? And is the result compelling in the sense of mankind? Or do the technologies create a reality that is logicaly and correct in the sense of a artificial intelligence? May be artificial intelligence develops its own value frame and own rules? Because artificial intelligence has to learn and develop independently in order to be...
Dear Petra,
you're welcome.
Although the book is not focused on your specific question, it provides a useful insights about the nature of models that use AI to analyze Big Data, and raise many important issues abot scientific research, justice, financial system and other fields.
I'm glad for your interest, since this problems became an important part of our education as citizens, due their huge impact on almost every aspect of our lives, from job market organizations and recruitment to basical infrastructures of sanitary systems, from transport to telecommunications ... in fact, the speed of change in the real world is many times greater than awareness of people about it.
Best regards,
Marco
Dear Marco,
your answer reflects my thoughts and also fears.
Best regards,
Petra
Dear Petra,
as for what concerns your specific question, an article on the web that I found very pleasant (and also well organized) is at the following link:
https://www.painscience.com/articles/bogus-citations.php
For an in-depth about falsifications in experiment findings an interesting lecture come from John Hopkins at:
https://www.annualreviews.org/doi/abs/10.1146/annurev-statistics-060116-054104?journalCode=statistics
In your question there are many points that could be faced, I found an interesting analysis in an article that you find attached in .pdf format.
We can think that, in principle, a well-built algorithm could help in prevent this phenomenon, if we maintain the awareness that the exact opposite could happen, and that the Data Analytics always requires an extreme caution in the evaluation of its results, due to the general problems I mentioned before.
As always, hard and patient scientific work is the only sure way to achieve a good quality work, and it is also the only way to evaluate it correctly.
I guess that the trade-off between quantity and quality of work will continue to exist, regardless of the increase in our technological possibilities.
Best regards,
Marco
Hi,
It is well known fact that motivations for citations are not always functional, that is, they do not serve the real purpose of building science layers. The more literature is published, the less relevant citations tend to be. There are ways to overcome this problem in bibliographic databases, but not in Google Scholar which is a search engine.
AI research will help solve these problems in the future, but in the time being, it seems reasonable to add experts' opinion to research evaluation. This could prevent somehow the deterioration in the quality of published research.
@Marco, dear Marco. Thanks, the article was very interesting. But it seems to me that he is primarly concerned with the problem uncited or seldom cited literature and less with false citations. Best regards, Petra
Petra Seidel Dearest Petra,
you're right, of course. Regard to the .pdf, In fact, the connection with your question is very indirect and I realize that I had not explained it at all.
The attached article is about of a problem that is complementary to that of false citations; while in the latter case non-existent or distorted citations are proposed, in the first case it's omitted to give the correct references.
I propose this work to your attention because the result of these two different problems produces a conceptually similar effect, that is literature that does not report its right references (where they exist).
Sorry for not having better explained the concept (indeed a little twisted :) ).
Marco
No problem, dear Marco, the paper was still interesting. And you;re right concerning the result of these two different problems. Best regards Petra
Marco Rao
dear Marco, I’m reading the recommended book of Cathy O’Neil now. It is really very exciting and informative ant the author explains well. In interesting other book was published this year: Frank Schätzing: “Die Tyrannei des Schmetterlings” (“The Tyranny of the Butterfly” but sorry, the book still it isn’t translated in English but I believe it will be)? https://www.dw.com/en/best-selling-author-frank-sch%C3%A4tzing-creates-an-ai-dystopia-in-latest-thriller/a-43506188
- Badges
- Science topic