How to do an effective literature review with Scholarcy

In our recent article: Five pitfalls to look out for when doing your literature review, we talked about some of the challenges researchers and students face when it comes to evaluating the articles that will be most for their own work. This can be a daunting task even for experienced researchers because of the sheer volume of literature out there.

There’s no shortage of discovery services available that can return pages of search results, even for obscure terms. But once you’ve collected all those papers that look relevant, how do you figure out which ones will be most useful to you, without attempting to read them all in full? The good news is that in recent years technology has really developed not only to help you find valuable research but also to digest and absorb that research more easily.

Generate summaries from a collection of papers to review more easily

There are a few different ways you can use Scholarcy to automatically generate summaries from a collection of papers. Whether you want to upload an entire folder of papers you’ve been saving over time or import a set of results from a recent search, this can be done really quickly by doing one of the following:

• Export search results from research platforms such as PubMed and import into Scholarcy.
• Create an RSS feed from a research platform using a keyword/search term so that Scholarcy continually pulls in any related articles and generates summaries from these in the background.
• Upload a folder full of papers from your Dropbox or Google Drive.
• Upload a.RIS file of references from your reference management tool such as EndNote to generate summary flashcards from a bibliography.

Turn your search results into article summaries

Today we’re going to focus on how Scholarcy can help you with your literature review by turning your results from search engines such as PubMed into a set of summary flashcards that you can read, share and annotate on any device.

We’ll take PubMed as an example but you can use this technique with any research platform that can export search results as a .ris or .nbib file.

1. First, run a simple search using, for example, the name of a drug or a medical condition eg. “Cancer immunotherapy”. Apply the necessary filters to your search such as publication date as well as selecting articles that have a free, full text version available.
2. Next, select the option of exporting your search results to a citation manager. This will create a .nbib file that includes the title, doi, author and abstract of every paper in your results which you can save to a folder.
3. Create a new folder in Scholarcy. You can name this using your search term to keep track of your summaries. A useful tip is to adjust your ‘View’ Settings (check the ‘Title’, ‘Author’, ‘Year’ and ‘Headline’ boxes and uncheck the others) so that the main finding from each paper imported will be immediately visible as you scan your summaries. This can really speed up the filtering process and get you focusing faster on the papers that will be most informative to your research.
   
4. Then you just need to import your search results file from Step 2 into your library. Scholarcy locates the full text of each of these articles and uses machine learning technology to distil each one into a summary flashcard. This may take a few minutes or more depending on the number of articles.

Tips for systematic screening now that you’ve generated your summaries

Whether you choose to import search results from a service such as PubMed; set up a direct RSS feed to automatically import the latest papers to Scholarcy; or upload a folder full of articles from Dropbox, the output is the same and there are a number of ways this distilled content can help you assimilate new and complex information without running the risk of missing something important.

Scholarcy summaries are structured in a consistent way to help you pull out and assess the most significant information from a paper more easily. They can also make key findings and results from similar papers easier to compare. So now that you’ve generated summaries from the longlist of articles you want to assess for your literature review, what’s the best way to systematically screen these?