It is totally unethical to cite a paper that you have not read personally, as that is how errors and misunderstandings are perpetuated. If you have not read it, then cite a secondary source, like a review.

The purpose for citing is to let readers know that a specific piece of information being provided has a source, other than author’s own observation or reasoning. In many cases, the strength and credibility of a work depends on the validity of its references, as well as the author’s ability to represent those sources clearly without plagiarizing. With this in background, it is immaterial whether a review article or original research is cited.

I think this is a very important question and needs detailed analysis. Suppose I am writing an original research article and for supporting my observation several other papers are to be cited. Now it is often possible that those papers are cited in some other review paper closely related to my work. In those cases, I think we need to include both. The review papers can be included in the introduction section, where general lay-out of the problem is presented. Original articles can be included in those sections of my paper where pin-pointed comparison or comments are required.

It depends on nature of reference. If author want to refer to a detailed work, original articles should be preferred. But if there is need to refer to only outline of work, review article should be preferred.

all research paper contains some specific information, so cite it where you need. all type of research papers are important (Letters, Reports, Review and even patents).

I favour referring to both the review article, and the original research. This can be particularly helpful to readers who may have limited access to journals, and if they cannot access one of the references, perhaps they can access another.

On the other hand one should be careful with citations given in review articles. E.g. one of my papers was wrongly cited in some big review - referred in section of alginate gels while being about pectin gels.

As authors we should be scrupulous. The best will be our work the higher notes we will get, and our article will be cited more often.

That is of course my personal opinion.

my opinion: review in Intro ...original in second part of intro and results part ... new original papers are still not placed in any review work ..

While I appreciate review articles, I try to reference original research articles. I feel this enhances the focus of the manuscript of which I am part and increases its relevance in the specific area. I will reference a review article if it has numerous original references that I utilize. The difficulty with review articles is verifying its validity as noted by Marta.

I think, both are important. While reading a paper with topic of our interest, we often look at the experimental procedures, recipes, etc. and also at the same time, look at evaluating articles and review papers that are referred to have better knowledge about the state-of-art researches going on, prospects of methods/materials used, etc.

Citation to the orginal research papers and earlier reviews are needed with regards to different aiming. Citations to the earlier reviews have the objective how the new review will cover deficiency and be a complementary of the knowledge to the earlier reviews. Citation to the orginal research papers both new and old one are needed in a review paper for further discussion. Of course, permission requires if the authors of planned review bring research results from other authors.

Both are useful. I tend to prefer to cite original papers for any specific information, whether or not there is a review. If I know of a good review I like to put something like "the field was reviewed by Who Ever et al." in the introduction. I actively avoid to cite reviews that I think are bad.

We can't bifurcate both review and research article. Both are very crucial for any researchers.

In my opinion, you have to cite those articles which related to your work. It could be review, letter or full length research article. If your work is entirely new then review paper is not a good option, then letter/communications would help. If you are discussing/reporting something which already exists then original research articles should be cited. Also, if you are discussing some general thing or outline of work, then cite review articles.

I think we should cite those papers which have reported that thing for first time, then you should not think whether it is a review or, research article or short note,even we should not think about the journal, if the report is published in average journal still it should get its due credit, many time people avoid giving credit to papers from small journals, i think we should respect the quality of work not the journals,

I also think that you should include both of the references about your previous reports and works of other research groups. These are mainly for comparison between different works: yours results and the other ones. I think that the reader understand better your work and your motivations, and he can comparate the differents results and extract his own conclusions or new ideas.

Considering the benefit of your future reader, both original articles and review are typically important. A colleague that is an expert in the field may know (or have written) some reviews of particular interest, so you may argue to concentrate on relevant original literature. However, I am very grateful to find a reference to a particularly helpful review when I read papers somewhat outside my main research focus.

Considering citing originial articles, I find it more important to really cite the relevant original article. Too often you find an article referenced for, e.g. preparation of a compound (in chemistry/materials science), but the cited article refers to another one, which refers to another one, which refers to another etc. In the end, you may or may not find an article described what you have been looking for.

That´s is not very nice, nor is it professional, in my opinion.

I believe there are two main aspects for citing references: one is to clarify the scientific mechanisms or views on the topics; another one is to give the detailed original development made by previous researchers. Ideally, if the two aspects can be combined into one, which means to cite original articles if possible. If it is not clear which article was the original contribution for particular model or fact, review articles are useful. Simply referring to a review article is not necessarily very helpful for readers in the case now researchers are easily to obtain a copy of review paper on a topic from scientific database.

I think the combination of both is important. I always try as much as possible to cite most relevant papers. People have put effort to clarify some scientific phenomena and citing their work is a complement to their effort. Of course, I have seen comment papers due to the neglect of a previous work in the same area. In essence, cite as many papers (whether review or PRL format) as possible that are relevant.

If you need to make a reference about the state of research in the research field (in particular, in the introduction of your paper), it is definitely advisable to use the review paper. In the case of discussion of original results, in my opinion, it is necessary make reference to the original papers. Therefore, you can use both review and original papers in different parts of your manuscript. Both reports are valuable. However, a review article becomes less valuable with time because more recent paper of this type appear. I think that the value of the original results did not decrease with time.

I fully agree with Yuriy comment, the order is important. It is also important to note that in certain specific areas of research review papers are rarely issued. Therefore, their real value is questionable. If an area of scientific research is very topical, current original research reports will not miss.

It is totally unethical to cite a paper that you have not read personally, as that is how errors and misunderstandings are perpetuated. If you have not read it, then cite a secondary source, like a review.

From my point of view, we should cite the papers which we have read and are original research papers. Giving reference without reading the paper thoroughly might be a blunder mistake sometimes. We should also cite the research papers of our own group related to the work so that the reviewer could observe our present as well as published work and its quality. It might help in the reviewing process.

I am sure that if you cited some paper during preparation your manuscript you must read their.

It really depends on the situation. If you want to cite a specific scientific advance, you should select the original paper. On the other hand, several review papers give you an overall view of the field of the research and often offer some perspectives on the current trends and future directions of that topic. This information can be very important for you to justify the rationale behind your work and in this case the review paper should be cited.

The first rule is to cite only the source from which you took the information.

If you have different sources of the same data, in describing concrete results it is better to give reference to the primary source, and in the analysis of the situation - to the recent review articles.

Agree with most of the responders, while a review paper should be used to link the readers who might want to know the state of the art in the particular research field, original research contributors must be citied while discussing/analyziing their contributions.

Any published peer reviewed article will do the job. But stay away from referencing original reports within Organization you work. These reports are not assecible over Internet, so, reference will be just ignored.

In my opinion, the original work can provide much more information about the experimental results. Maybe these information have not been discussed in this publication, but maybe it is useful for us. Therefore, I think the review paper can help us to understand the overall area, but the original work is much more helpful.

Review paper can be narrative or systematic. If you need selected and specific data from any of this two kinds review paper we need use original references. Also, if you use opinion of author from systematic review, I suggested that use this reference for cite. Simply, which results or conclusion or opinion use - you need cite them.

There is no real problem. Both, review article and original articles can be referenced within a new pulication.

Maybe this is not strictly related to the question, however the San Francisco Declaration of Research Assessment (DORA) (see http://am.ascb.org/dora/) makes mention to how researchers should choose the refferences:

" 16. Wherever appropriate, cite primary literature in which observations are first reported rather than reviews in order to give credit where credit is due."

The question to me is related to this paragraph. The issue is about giving credit to the researchers that deserve it. Writing good reviews is not trivial but it is probably easier to do ground-breaking research.

I hope this helps with this interesting discusion.

" 16. Wherever appropriate, cite primary literature in which observations are first reported rather than reviews in order to give credit where credit is due."

I totally disagree with DORA. It says nothing about actually having to read the primary sources, which is essential to avoid the constant misrepresentation, half-truths and errors that are perpetuated by reliance on second-hand sources. It is also better to be credited with some good idea from a secondary than primary source. For example, I think Wells was the first to come up with the concepts of natural selection, genetic drift, and the out-of-Africa human origin. The actual reference for this is quite obscure, and very few are going to bother to read it. Furthermore, if they did there is some chance they might conclude that I have misunderstood or overinterpreted what Wells wrote, diminishing the reputations of both Wells and myself.

I agree with Anthony Gordon and with those who share his opinion. Cite primary literature whenever possible. However, when that literature is more than 10 years old, it is reasonable to cite a review, BUT I like the citation ("Reviewed in (X)). For some groundbreaking papers, I continue to cite the original paper, even beyond 10 years. I think that it is particularly important that students and fellows learn to read the primary literature and not take things as fact just because they are cited in reviews.

Some journals limit the number of references in a paper, and this is where one has to sometimes think about using reviews even if the primary literature is still recent.

Yes, it is important to cite new review published in SCI journals, however the original work which has reported first also need to be cited

"the original work which has reported first also need to be cited"

First, it needs to be read.

I think, it is always better to cite the original work unless otherwise they are not available, in such case citing survey is well considerable.

Another common problem I have noticed in novice writers is to cite a nonreview paper as if it were a review.  For example, mention of primary data in the introduction of discussion of a nonreview paper is cited, instead of finding and reading and citing that original paper.  This should essentially never be done, but some beginning writers do not know this.

"mention of primary data in the introduction of discussion of a nonreview paper is cited, instead of finding and reading and citing that original paper.  This should essentially never be done, but some beginning writers do not know this"

The reason for this is that even the top journals do not insist that every paper cited has actually been read.

I always try to cite the original works, but sometimes I have a problem to obtain a source journal. In this case I cite an other relevant article published at approximately the same time.

"In this case I cite an other relevant article published at approximately the same time."

I agree it is often hard to obtain the original source, but I still say you should only cite articles you have actually read (and not just the abstracts).

You need to cite the article you have read before your research. Try to collect the research papers concerned to get the real situations and environment of the results concerned. If they are not available at all, then you have to cite the review paper from which you got the concepts and results concerned.

I think Anthony Gordon hits the nail perfectly: the original work needs to be read. AND understood. AND summarized in an ENLIGHTENING way. (Sorry about the shouting.)

I really do not like reading a paper whose only comment on the subject prior to the paper you are reading is something like: "for recent work in the field, see [1, 4, 5, 9, 11-23]. How utterly boring can it get!? Such a practice is extremely lazy, and it serves no purpose at all except to fulfill the wish of (some) journals to "cite recent papers". 

If it is difficult to locate an original paper, there is a very good methodology: check with your local librarian, and ask for help. Whether it takes a week or two to find the paper does not matter - you need to get hold of the original source, so that you can read it first-hand. Only citing others who have read it leads to the risk that the paper in question will eventually be cited in the wrong way - that is exactly like that game of whispering a message from pupil to pupil across the whole class-room, with the inevitable effect that the original message will be heavily distorted (and shrunk). Do NOT make that mistake - find the original source!

Somewhat of a re-iteration, perhaps:

If a journal limits the number of papers that you may cite - and you find that your paper would be incomplete without adding more (with proper referencing of their relevance to the research you perform in your paper, of course) - then DON'T publish in that journal. If a publisher bans long reference lists then their motive is to cram as many papers into one issue as possible, and I deem such journals unserious. Don't fall into that trap, and publish elsewhere.

"If a publisher bans long reference lists then their motive is to cram as many papers into one issue as possible, and I deem such journals unserious."

Popular journals are popular for a reason, and so are limited for space.  Would you want to publish in a subscription journal that has plenty of space for your work? 

In my opinion, Both reports are valuable. However, a review article becomes less valuable with time because more recent paper of this type appear.

Hello all, currently I'm preparing a literature review as part of my final thesis. To do so, I really would like to use only peer-reviewed articles but do have a lot more results for research paper by searching in well-known databases like Sciencedirect, Emerald etc. Do you think I can start working with research paper or would you rather stay with pee-reviewed articels only?

Dear Nadine, I think this depends on priorities issues related to your research. Whether the time is enough for you to head for research papers, then it is fine. This will be much more better, as the huge no. of related research papers that you have read will be classified based on the research topic that you determined for your paper. However, if the time is catching up with you and you are requested to finish up thesis writing, then you need to write your literature review in the well-known traditional fashion by assimilating all related researches into it. Later, you can write your research articles based on the review written.

Regards

Dear Moayad,

thank you very much for your answer. That's a good point.

Best regards,

Nadine