6

Problem

I have to familiarize myself with a new sub-field fast, to the point where I can recommend an agenda/recommendation for further research. I thought about doing a systematic literature review to achieve this, but have to work somewhat cost-efficient due to external deadlines that I have no influence on.

Background

Systematic Literature Review

When I say "systematic literature review" (SLR) I'm referring to a speficic concept which originated in medical science and was introduced to computer science by Kitchenham (2004), for a hands-on guide see Kuhrmann et al. (2017).

Purpose

The purpose of conducting a systematic review would be to acquire the necessary knowledge to be able to argument for new research endeavors (i.e. defend their relevance). Additionally, it would be nice to get a publication out of this, which is part of the reason why I think about doing a systematic review instead of just writing up a state-of-the-art document for myself.

Narrowness

The topic I want to research is part of Software Engineering, and deals with the application of a specific concept in a more-or-less specific environment. I could not find any systematic reviews (or indeed, any reviews) covering this.

What I found

Ressources

Time

The approximate time to become familiar with the state-of-the-art in a field that one is already familiar with, according to this answer, is one year. Even though the question was about a PhD, and not just a work project, this strikes me as a little high. An answer to a question about the usefulness of SLRs warns that spending 6 months on a SLR for just a paper is excessive.

Human effort

Even though the literature on SLR recommends multiple researchers, this answer says that it is quite possible to conduct and report a SLR on your own. I might get a senior colleague to look over my draft and maybe recommend specific papers, but I should expect to do almost all of the work on my own.

Paywalls

I don't have institutional access to a lot of journals and such, which means that I will mostly be able to read exactly those papers that are either published in Open Access journals or are otherwise available through researchgate and similar platforms. This answer predicts that a restriction like that will render the SLR of little interest to publishers. Since I not only support Open Access publications for various reasons, but also really don't have a lot of choice in the matter (I can't convince my institution to pay for the access I want, nor do I have the money to pay for all those subscriptions), I do hope that this is not true.

Usefulness

Even though it is considered useful to write a SLR in order to get a good foundation in a field, it might be more efficient to just read some papers (specifically reviews, if they exist) and follow the references. However, this efficiency calculation does not factor in the expected utility of publishing a SLR as paper.

Questions

Main question

Is it really useful to write SLRs in order to familiarize yourself with a topic in a short time, while a publication is nice-to-have but not mandatory?

Sub-questions

  1. Is it possible to do a SLR on your own, in a matter of weeks?
  2. Do paywalls effectively keep me from writing and/or publishing SLRs?
  • 2
    Please read Literature searches in publications when you have limited access to journals. I am an independent researcher, have no institutional access. This Q&A has helped me a lot. – scaaahu Sep 2 at 13:53
  • 2
    Yes, paywalss effectively keep you from publishing an SLR. "I could not find or read that relevant article because of a paywall" is absolutely not an acceptable excuse to fail to include relevant literature. Most kinds of research cost some money for research resources or data access; SLRs require access to journals as their critical data source. Either your institution pays or you pay. – Tripartio Sep 2 at 15:56
  • Your question is hard to answer because they are three distinct questions, each with a distinct answer; this is against the Academia SE rules, where each question should ask only one question. Even though the background is common and it would be repetitive, I recommend that you repost this as three distinct questions. I might be mistaken, though, in my assessment. – Tripartio Sep 2 at 15:59
  • @Tripartio thank you for your feedback. I have edited the question to reflect the fact that there is one main question, with two additional questions clarifying specific aspects of the main question. I believe they are more useful as sub-questions then if I'd ask them on their own. – nonthevisor Sep 3 at 8:07
  • I won't say that sub-question 2 is effectively solved by the existence of SciHub. Such a statement would not be in the interest of publishing companies, a type of business that relies on the efforts of tax-paid researchers without paying them, leading to huge profits. – lighthouse keeper Sep 3 at 9:50
1

Performing a systematic review is very time consuming. Also, it may result in a "high dispersive work", if a very specific research question is not set "a priori". Also, SR is not only "indexing" evidence, but also effectively aggregating them in a readable and useful summary. If you are not an expert in that field, you may not be able to do that, and the quality of the overall work may be low.

If your goal is to improve your knowledge in a specific field, performing a systematic review may not be the most suitable option in the first place. You'd rather read existing reviews and summary (if any) - this would be extremely more time-effective.

To answer to the other two questions:

  1. Yes, it is possible to do in a matter of weeks, but not entirely on your own (the process of screening requires at least two independent reviewers, according to the PRISMA guidelines).
  2. Yes, if you are not able to reach those literature that meets the inclusion criteria of your SR (and this is quite probable).
| improve this answer | |
0

Systematic literature reviews should surely be written by domain experts, since they have a better understanding of the broader literature. Attempting to write a review to familiarize [one]self with a topic seems rash, studying existing reviews would surely be a better approach. Alternatively, studying key journal articles and textbooks is an option. The former should include related work sections which are essentially mini literature reviews.


The topic I want to research is part of Software Engineering...I could not find any systematic reviews (or indeed, any reviews) covering this.

Software Engineering is an established research field. Numerous systematic reviews have been conducted. Books have even been written. I'm struggling to believe nothing exists, especially as related work sections can be a substitute for systematic literature reviews.

| improve this answer | |
  • 4
    As a software engineering researcher, I can name 10 or 20 topics for which no systematic literature reviews exist off the top of my head. We're faster at coming up with new topics than we are at aggregating and systematizing knowledge. – lighthouse keeper Sep 3 at 9:43
  • @lighthousekeeper So numerous topics have reviews/books. My point is: Established topics have systematic literature reviews. For new topics, the entirety of the literature can be perused. – user2768 Sep 3 at 10:20

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.