16

In disciplines such as computer science and engineering, Google Scholar is a fundamental tool in not only searching for recent work of a researcher but is also frequently used as a metric for the proficiency of the researcher.

I've noticed in many other fields, e.g. political science, or even math, Google Scholar is not nearly as popular.

Does anyone have any insight into why this is the case? Do these other fields have other tools they use to follow researchers or measure their output? Perhaps these factors are not as important in these other fields?

  • I should comment that Google Scholar is a problematic tool to use, since Google tracks people across its various services, building records and profiles, and using them not just for personally-customized advertising (which is bad enough IMO), but also copies all of this aggregated data to the US government. So, to be specific: Your NSA records now also include which articles you're looking for and choose in the search results. – einpoklum Jan 1 at 21:36
21

Math already has an excellent tool, MathSciNet. And, more importantly, MathSciNet is much more accurate than Google Scholar.

If you are a mathematician, look up yourself in both places and compare!

  • 1
    It’s a trade-off though, MSN only includes published papers and not preprints, so it’s more accurate but less extensive. Also google has features like searching within papers that cite a given paper. – Noah Snyder Dec 31 '18 at 17:58
  • 7
    +1 This is a useful answer. However, I think calling Google Scholar "less accurate" is not exactly correct. I think what you mean is that GS is more inclusive in indexing. So GS counts citations from arxiv preprints and the like, which some people think should not be counted. MathSciNet is more exclusive. It's not a question of accuracy, they measure different things. – Thomas Dec 31 '18 at 19:18
  • 2
    Similarly, astronomy/astrophysics has the Astrophysics Data System and has had it for a very long time: ui.adsabs.harvard.edu. This is also more accurate, complete, and up-to-date than Google Scholar. It also indexes preprints. – NeutronStar Dec 31 '18 at 19:32
  • 1
    Even in computer science, if I'm looking for a particular paper or papers by a particular author, I usually try DBLP before Google Scholar. Being based on manually-curated conference and journal metadata, DBLP's coverage isn't as broad but the quality of results is much higher. – Jeffrey Bosboom Jan 1 at 2:16
  • 1
    @Thomas Not to mention that GS will give a more up to date picture than MathSciNet, since the latter can take quite a while to actually index papers sometimes. – Tobias Kildetoft Jan 1 at 13:23
10

In computing science, papers are reviewed rather quickly (because they are often tied to conferences, with deadlines, and without journal backlog). They are also usually indexable.

In biblical studies, to give another example, it can easily take over two years from submission to publication, even with a minor revisions review shortly after submission - just because journals have a large backlog. Also, there still are journals without digital edition. Biblical scholars I know rely on mailing lists for tables of contents and find PDFs of articles they want to read on academia.edu. Google Scholar simply does not index everything and is not fast enough.

From what I've heard, in math it's the review cycle that takes long. This is different from a journal backlog, but the result is the same. Arxiv will be more up to date than Google Scholar.

  • 9
    But Google Scholar indexes arxiv and even things like a pdf on an academic's homepage. So I don't think this entirely explains it. – Thomas Dec 31 '18 at 10:06
  • @Thomas it only adds PDFs from personal pages or things like academia.edu when it has already indexed the article from a reliable source. I can't speak for arxive, but I know that several articles I had on academia.edu were not on Google Scholar until they were officially published. – Keelan Dec 31 '18 at 10:13
  • 11
    I definitely have papers indexed by GS that are only on arxiv. – Thomas Dec 31 '18 at 10:25
  • 1
    "Arxiv will be more up to date than Google Scholar." Arxiv is a repository of pre-prints; Google Scholar is an index of publications. So I'm not sure what you mean by this apples-to-oranges comparison. Also, as a computer scientist, I dispute your claim that CS papers are "usually" in conferences. – David Richerby Dec 31 '18 at 23:04
  • 1
    @DavidRicherby Since time to publication is extremely long in mathematics, and since essentially everyone posts on arXiv before submitting for publication, arXiv is functionally equivalent to an index of publications. Furthermore, most journals now allow leaving the arXiv posting up even after publication. – Elizabeth Henning Jan 1 at 2:06
7

There is no single answer to this question. Every field is unique. There is also an element of randomness: If a handful of the right people create Google Scholar profiles, others will follow and adoption will rapidly spread. However, I think there are two key factors:

  1. Willingness to adopt new technology. This depends on the existing alternatives (e.g., MathSciNet) and also how tech-savvy people are. I've noticed that, even at the individual level, there is a strong correlation between having a good, up-to-date website and having a Google Scholar profile.
  2. Publishing attitudes. Google Scholar is very inclusive in indexing. It counts citations from arxiv preprints, not just peer-reviewed journals. I even get citations from lecture notes and research statements that people have uploaded to their websites. Different fields have different attitudes about what should "count" as a citation and that will affect how they view Google Scholar. For example, math is much more conservative than computer science and Google Scholar adoption reflects that.

A final note specific to computer science: Just about every computer science researcher knows people (e.g. former students) who now work at Google. And Google also produces a lot of computer science research. Thus computer science has a very positive view of Google, which may not be the case in other fields.

6

"Time to market" of the paper (e.g., shorter review time) as Keelan noted might be a reason.

It might be also related to what is indexed by google scholar vs. more official indexes. (see link below). For example, google scholar counts as a citation basically everything that looks like a citation, regardless of whether it was peer-reviewed or not. It is possible, that some fields are more reluctant adopting such metrics. For example, arXiv platform has been long ago a popular in math field. But those papers are not peer-reviewed, at least for posting them, there is no requirement of peer-review. In other domains (I think biology) preprints have become popular more recently.

Another reason: in CS there are a lot of people outside academia. They are interested whether someone relevant cited their paper (or whatever it was). This boosts google scholar use. In many other fields researchers work only in academia. the official representatives might tend to rely more on official indices (e.g., for promotion etc).

It is also possible that some fields are more conservative than others. CS is in general a relatively young field. Also, it is more natural for CS to adopt new technologies. But think about philosophy field that has longer traditions. People in such fields also less care about technological innovations.

https://www.researchgate.net/publication/239314956_Citation_Analysis_Comparison_of_Web_of_ScienceR_Scopus_SciFinderR_and_Google_Scholar

-8

Probably there are more computer-illiterate people in these fields.

  • 2
    @Buzz: This is clearly not a valid comment; it does nothing comments are designed for. You may still consider it wrong, lacking explanation, or similar, but it clearly belongs in the answer domain. – Wrzlprmft Dec 31 '18 at 13:46
  • 7
    This would imply that math has sufficiently many computer-illiterate people that they cannot use Google Scholar effectively. Given that mathematicians regularly use LaTeX and put up preprints on arxiv this seems unlikely. – JoshuaZ Dec 31 '18 at 14:15
  • 4
    Google scholar is easier to use than other indexing services. It's basically made for computer illiterates. – henning Dec 31 '18 at 18:43
  • 1
    @DanielR.Collins while different people's experience is, of course, different, I am yet to come across someone like you describe. Even the pretty elderly people in my department all use email, and almost all of them (that I have interacted with) use LaTeX. In fact, I can only think of one example of someone producing some maths stuff (that I have personally seen) that wasn't in LaTeX. Given that journals ask for tex files, as does arxiv, I personally expect the use is actually quite universal. Only my experience though! – Sam T Dec 31 '18 at 18:59
  • 2
    I disagree, @Buzz This is an answer to the question. It's not a good one, but it is an answer and there is probably some nugget of truth to it. – Thomas Dec 31 '18 at 19:11

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.