Skip to Main Content
It looks like you're using Internet Explorer 11 or older. This website works best with modern browsers such as the latest versions of Chrome, Firefox, Safari, and Edge. If you continue with this browser, you may see unexpected results.
Lund University

Writing and sources

A guide on writing and the use of sources intended for students at the department of Biologi and CEC.

Evaluating scientific sources

Scientific sources should contain reliable, peer-reviewed and research-based information. You should always evaluate the source yourself to make sure that it really is scientific, even if you find it thru a trusted channel. This becomes even more important when the source is found through an informal channel, for example a website or a search engine. Here follows some general instructions for how to determine whether or not a source is scientific. For additional instructions concerning how to evaluate journal articles, see section Scientific articles below. 

When evaluating a source, start by finding out if it has been peer review. Scientific sources should always have gone thru some type of peer review, i.e. a process where the work has been examined and evaluated by experts prior to publication to ensure that it is of good quality. When an article is submitted to a scientific journal, the peer review is carried out by other researchers in the same field (the author's peers). The peer-review process for books varies between publishers. Some will have a rigorous peer-review process similar to the one for scientific articles. Other publishers leave the peer-reviewing to the editor, who may or may not be an expert on the subject in question. 

The next step is to evaluate the author(s) of the source. Who are the authors and are they experts within the field? What else have they published? Do they come from an established university? In addition, you need to evaluate the provider of the source to make sure that there is an established and well known academic publisher behind the publication. On the publisher websites, you should also be able to find information on the type of peer-review process they use. 

One common way of evaluating sources is to count the number of times they have been cited by other scientific publications. A large number of citations does however not necessarily mean high quality. Publications dealing with controversial subjects or studies that are considered bad science may also generate a large number of citations regardless of their quality. The opposite applies to publications with few citations, which are not necessarily of bad quality. Publications that are new or deal with narrow subject might have a low number of citations due to lack of time or opportunities to be cited, despite being of excellent quality. In addition, what counts as many or few citations depend on the type of material you are evaluating, articles are for example cited more times than books on average and the research subject.  

An alternative quality measure which is becoming more common is altmetrics, that is counting the number of e.g. social media shares, blog posts, downloads or use in reference tools. However, altmetrics suffer from the same problem as citations, namely that a publication can score high for negative reasons.

Scientific articles

A scientific articles should always be peer-reviewed. Sometimes it is stated in the article that it has gone thru peer-review. If not, read about the journal's peer-review process on their website. 

Scientific article are published in academic journals. If you are unsure about whether a journal is academic or not, look for information about the journal’s audience, purpose and review process on the publisher’s website. You can also check if the journal is listed in trusted databases such as Web of Science (Journal Citation Reports), Scopus (Sources) or PubMed. Note however that a journal could be absent from these lists if it is new and has not had time to establish itself. For articles in open access (OA) journals, make sure that it is a serious (non-predatory) journal by checking if the journal or publisher appears in online lists of serious OA journals and publishers such as the Directory of Open Access Journals (DOAJ) and the Open Access Scholarly Publishers Association (OASPA). Another common way of evaluating journals is to compare the average number of times the articles in each journal have been cited by other scientific publications, their so called impact factor (IF). Journals that are new or deal with narrow subject might have a lower IF due to lack of time or opportunities for their articles to be cited. In addition, what counts as high or low IF will depend on the research subject in question. 

The article should be written by and for researchers. Make sure that the author's affiliation is clearly stated, and point to a recognized research institution. The language should be scientific, using established terminology. Furthermore, scientific articles are, at least in Science, usually written in English. Scientific articles typically follow the structure Introduction-Methods-Results-and-Discussion (IMRaD), or a similar structure (i.e. Introduction-Body-Conclusion). Most importantly, you should be able to follow and verify the research methods used and how the results where arrived at. Last, but not least, the article should refer to other scientific sources, with both in-text citations and a reference list at the end of the document. 

Evaluate your scientific articles using the checklist below. 

Checklist for articles

Check if the article is peer reviewed and if it is available in any of the large academic databases (Web of Science, ScopusPubMed etc.).

Check if the journal is indexed Web of Science (Journal Citation Reports) or Scopus (Sources).   

For OA journals, check if the journal or publisher appears in online lists of serious OA journals and publishers such as the Directory of Open Access Journals (DOAJ) and the Open Access Scholarly Publishers Association (OASPA).

Evaluate the content of the article:

  • Method - Reasonable experimental design, no. of replicates and use of statistics? 
  • Results - Are all results presented (i.e. no hiding of negative or unexpected results)? Are tables and figures understandable? 
  • Discussion - Reasonable interpretation of results?
  • References - Number, type, quality and relevance?
  • Language/readability - Has the copy editor done his/her job? Bad copy-editing doesn’t have to be bad research, but warning sign!
  • Design/layout - Does it look professional? Bad design/layout doesn’t have to be bad research, but warning sign!