I'm about to submit my paper which is about predicting the apparent geothermal gradient in a country using a machine learning algorythm, the main motivation is the fact that the country only has heat flow data for about half of the territory and it's measured in-situ from well logs, so there are a considerable part of the country devoid of data. I'm using a Machine Learning approach to estimate the geothermal gradient using geophysical and geological data available for all the territory.

I referenced 18 papers: data sources, XGBoost fundamentals, decision trees, geological information about the country, current geothermal gradients map; and some about similar jobs (Predicting heat flow in Greenland, predicting heat flow in Antarctica, Predicting very shallow geothermal potential in Switzerland, Statistical modelling for global heat flow, etc).

So, should I add more citations? The Antarctica paper has over 60 citations, but they are doing double job (They train and test the model in Australia and then use the model to do predictions Antarctica).

I'm not quite sure if 18 references is fine for my paper, as it not a survey research or a review, just a proposed methodology. Thanks in advance for any suggestion!

  • 10
    There is no rule for the number of references in a paper. You should cite the papers you know about that are sufficiently related to what you are writing about. If that's what you did, then you're fine.
    – Dan Romik
    Dec 3, 2023 at 18:21
  • 2
    This question very much seems outside our scope to me, as it's about the content of research.
    – user176372
    Dec 4, 2023 at 1:53
  • @user176372 this seems an OK question to me - where else would you suggest asking it?
    – user126108
    Dec 4, 2023 at 2:52
  • 1
    What does your adviser say? Dec 4, 2023 at 4:09
  • 2
    @user176372 In my opinion, OP does go into unnecessary detail about the content of their research, but the underlying question is essentially "how few citations are too few in a research paper".
    – Anyon
    Dec 4, 2023 at 13:22

1 Answer 1


There is no "right" number of citations for a paper, only the number that you need in order to appropriately position the paper with respect to its background.

There is especially little correlation between the length of a paper and the number of references that it has. That is because it in papers with new results, most of the references tend to be in the introduction, while most of the length of the paper is typically in the methods and results.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .