Can I publish a survey-paper on a broader subject of a technology instead of narrow approach?

manpreet Best Answer 2 years ago

Can I write a survey on a 'Broader' subject of a technology, discussing how literature has discussed the technology, instead of a particular aspect?

For example in 'cellular communication for vehicles' there can be a number of ways to explore the field, including 'implementation challenges', 'Resource allocation', 'security', 'coding', 'vehicle-to-infrastructure communication'...and the list goes on.

Is it necessary that I present the survey on any one of this narrowed field. During my Ph.D. I have read a lot about almost all of these topics, can I combine them into a survey on the lines of "Survey of communication technology in Vehicles" etc. which details all these aspects?

manpreet 2 years ago

In theory? Sure, you can write a survey paper about any topic you like, no matter how broad.

In practice, there are two main restrictions: practicality, which is - how the heck will you write a proper review of a subject so large given resources that are available to you, and is now the right time for you to do that?; and publishability, which is your ability to find an appropriate venue that considers it on topic and get reviewers to sign off on it as good.

For the obstacle of practicality, as your starting scope becomes broader more papers must be included, until it quickly becomes unwieldy. Using your example, a search on Scopus for "communication technology" AND "vehicles" returns 265 articles in 2017 alone, and only slightly lower amounts for previous years. Removing the word "technology" shoots the results into the thousands per year, as an example. Clearly not all of the papers returned by a keyword search such as this are relevant (and not all relevant papers are turned in 1 keyword search), but a proper review paper covering even only the last 5-10 years would still be a considerable undertaking. The length of time you need to cover depends on the publications in your field, and I've seen some reviews cover 50 years, which obviously would not work if the scope included 10,000+ papers.

This leads into the issue of publishability. One problem is by being more broad, some of the more specialist topic venues might consider it out of bounds. But honestly, the more likely problem is that as you broaden the topic, it becomes harder and harder to do a good, convincing job of covering so much material. The exact numbers vary by fields, but in fields I'm familiar with a published survey paper generally covers at least 10 years of work, and is sufficiently bounded in scope to include around 200-300 papers. Re-using the example from Scopus, if even 20% of the papers returned in a keyword search were relevant you'd already be at over 400 papers already, which only includes searching in title/abstract/keywords. I've also been repeatedly told this sort of work is hard to publish at the top venues my colleagues care about, but this of course varies by field (its also rare to publish meta-analyses in my field, but this is certainly not true in other fields).

Reviewers will also inevitably ask, "what is new here?", so you'll need to explain how such a review covers things that no one else has already covered in this way. If 3 papers review A, B, and C, and you propose A + B + C, it will tend to be seen as "nothing new, reject" - and if you make it A + B + C + D, only D is new, which can also encourage rejection because only a small percentage of your review is 'new'.

The other big issue - which draws on both these restrictions - is what the "point" of the paper might end up being? If its just "here's a bunch of papers", that is often seen as less interesting than if your review also systematically addresses interesting research questions. So if you, for instance, set the scope to be "challenges for industry adoption of between-vehicle communication technologies", your systematic review is more structured, it will be less likely to read as a paper-dump, and conveniently in your selection of papers you will have more criteria to look for that allows you to throw papers out as unrelated - narrowing down the papers you need to actually cover in depth. But if its just "lets talk about a bunch of papers that happen to be about this general topic", you are going to need a lot of work to explain what is "interesting", and a reader might need to put in more work to figure out why this is worthy of publication - which generally does not bode well for the project.

None of this results in a "yes, do it" or "no, don't do it". As always, the best two sources of guidance is the literature in your field (what are published surveys, how are they structured, what do they cover, where are they published) and senior researchers who may have been involved with surveys before and can give you first hand experience more relevant to your specific field.