Descriptive Transcript

When visual content is read or explained in audio, it is not reproduced in the visual column below.

Audio Visual
Welcome to Deduplicating in Rayyan, a tutorial from Northeastern University libraries. Deduplicating in Rayyan Title Slide
In this tutorial, you'll learn how to use Rayyan's deduplication tool to remove any identical citations that may have been found in your database searches. This is a common occurrence, so deduplication should be part of all evidence synthesis workflows. Learning Objectives: You will learn how to use Rayyan’s deduplication tool to remove any identical citations retrieved from database searches.
At this point, you have already created your review in Rayyan and imported the results of all your searches. Remember, you shouldn't deduplicate until all your search results have been imported. Otherwise you will have to deduplicate again. Your initial pool of results will include a significant number of duplicates. You want to remove duplicates before beginning the screening process. It is useful to understand at the outset Rayyan's deduplication tool is mostly manual. Duplicate records will not be automatically removed. The deduplication process is partially automatic and partially manual. Rayyan automatically identifies possible duplicates and groups possibly duplicate records together. Your team manually confirms whether or not records are duplicates and deletes true duplicates.  
The first step is to click on the button, detect duplicates. Once duplicate detection has completed, there will be a new box in the left panel: duplicates. Step two is to select unresolved from the duplicates box so as to limit to only those unresolved duplicates. We can see now that unresolved is highlighted. You can also see text which reads, showing one to six of four hundred ninety-nine unique entries filtered from seven hundred and eight total unique entries. Rayyan review page. Detect duplicates button appears in main panel below title. Selecting unresolved from the duplicates box filters the entries in the main panel below the “showing # entries” alert text.
Now we see stacked paper icons. This denotes that there are multiple, possibly duplicate, records of this specific article. We can click on the first article here and now we see the article's title and abstract in the lower pane. Step three is to select, resolve duplicate. Zoomed in view of entries from main panel. Resolve duplicate button appears in lower pane after selecting a record.
After selecting resolve duplicate, you will be able to view the records related to this article, records which Rayyan identified as possibly duplicate. Step four is to examine the confidence percentages given to each record. This describes how confident Rayyan is that this record is a true duplicate. Next to the confidence percentage, you'll notice a delete button. When the confidence percentage is ninety five percent or higher, you can confidently delete the access record. After deleting the duplicate, we are left with one record. Now we can move on to the next set of identifying duplicates. In this case, confidence percentages are below percent. We want to examine these more closely. By comparing article citation information, the author's, journal name, volume and issue number, pages and publication date, we can easily determine that these are not duplicates. They are distinctly different articles. So in this case, we will want to flag these articles as not duplicates by clicking on the button labeled not duplicate. Zoomed in view of lower pane with possible duplicate records. Pane shows article information and deduplication buttons.
Thanks for watching. Closing Slide