Skip to Main Content

How to read a scientific paper

Tips to learn the process of reading and evaluating a scientific paper.

Types of Research Articles

The major types of research articles:

  • Manuscript/Research article: A formal research article that is submitted to the editor of a journal for publication, peer-review and/or to a preprint repository.  It reports the results of original research and assesses its contribution to the body of knowledge in a given area.  See "Format of Research Articles" below.  Considered a primary source article.
  • Review article: An article that synthesizes and evaluates information from several primary articles.  It does not describe original research conducted by the author(s). Instead, it gives an overview of a specific subject by examining previously published studies on the topic. By interpreting the findings of previous studies, review articles are able to present the current knowledge and understanding of a specific topic.
  • Peer-reviewed article: A manuscript or review article that has gone through an evaluation process in which independent expert scholars critically assess the quality and scientific merit of the article and its research.
  • Preprint: An article that has not yet been peer-reviewed. Repositories like arXiv, bioRxiv and medRxiv house pre-print articles to allow for immediate access to research. It is the responsibility of the user of such preprints to vet the information in them and to regularly monitor the preprint repository to follow the status of the preprint.
  • Case Report: Typically used in medical related fields.  It is a detailed report of the diagnosis, treatment, response to treatment, and follow-up after treatment of an individual patient or up to several patients.  It is used to describe a rare or unique condition or unexpected associations or findings.
  • Conference Proceedings: In certain disciplines like Computer Science, Mathematics, and some sub disciplines of Cognitive Science, and Physics, conference proceedings are peer reviewed by conference committees and are considered equal to the process of peer review by a journal. This can be confusing since in other disciplines conference proceedings are not used or cited in scholarly wiring. When in doubt ask your professor for clarity.
  • Dissertations and theses: Dissertations and theses are reviewed by committee members of the authors and while they are useful in research, most science scholars will recommend using the peer-reviewed papers published from these documents rather than citing the document itself. Again, there are disciplinary differences and when in doubt, reach out to your professor for clarity.

Format of Research Articles

Typical format of a research article:

  • Abstract: Concise summary of the whole article that includes information from the introduction, methods, results, and discussion.  Many times will have a "keywords" list at the end.

  • Introduction:  Introduces the topic and explains the purpose of the study and its importance.  Typically includes the hypothesis.

  • Methodology/Materials and Methods: Specific details of how the study was performed.

  • Results:  The findings from the study.  Usually includes data, tables, charts, and graphs.

  • Discussion:  The authors' analysis and interpretation of the results and significance.

  • Conclusion: Final thoughts and conclusions.  May include strengths or weaknesses of the study and suggestions for future research.  Sometimes this is combined with the Discussion.

Journal, Author, and Article Metrics

  • Journal Impact Factor:  A ratio which divides a journal’s received citations by a count of its published articles. The impact factor measures the impact of the journal, not the contribution of individual papers or authors.  The impact factor should not be used alone in assessing the quality of a journal.  Differences in the average number of references for an article can influence citation rates. 

  • Journal Citation Indicator (JCI):  An attempt to "normalize" the Journal Impact Factor.  Because it is normalized, it allows for comparisons across disciplines.  A JCI of 1.0 is average, higher than 1.0 is a higher than average citation rate, and below 1.0 is a lower than average citation rate.  This metric is freely available on the Web of Science site, unlike Journal Impact Factor which requires a specific subscription.  

  • H-index: The h-index attempts to measure both the productivity and impact of an author. H is the number of articles published by an author which have each been cited at least h times.  An h-index of 20 signifies that a scientist has published 20 articles each of which has been cited at least 20 times. 

  • Citation number ("times cited"):  The number of times an article has been cited in another published work.  Recently published work may have low citation numbers.  Also note that review articles are cited disproportionately, author's cite themselves to raise their citation counts, articles in certain journals are more likely to be cited than others, and, of course, people may cite an article because they are responding to errors or other problems with the article, not because it represents excellent scholarship.

  • Altmetrics: A way to measure and monitor the reach and impact of scholarship and research through online interactions. Altmetrics stands for "alternative metrics." The "alternative" part references traditional measurements of academic success such as citation counts, journal impact factor, and author H-index. Altmetrics are meant to complement, not totally replace, these traditional measures.

Biases in Metrics

It is important to look deeper than just the metrics defined above when thinking about the value or contribution of a scientist or a lab. There are on-going debates about the biases that publication metrics bring into promotion pathways, funding priorities, etc. Here are a few articles that talk about different perspectives to be aware of while thinking about scholarly metrics.