How to Read an iThenticate Report [Includes Example Report]
Learn how to read an iThenticate report! Find out how they work and how much similar text is allowed in a manuscript.
Updated on February 11, 2019
What is an iThenticate report?
Plagiarism detections tools are commonly used by academic journals to protect themselves and the academic community from the publication of plagiarized text. One of the more commonly used tools is iThenticate (also known as CrossCheck or Similarity Check).
How does iThenticate work?
In brief, iThenticate compares a manuscript to their proprietary CrossRef database, which contains a large number (tens of millions and counting) of documents from scientific conferences, journals, and
books. The tool will also do a search of the Internet and a search of the databases of several other content providers.
The program then generates a Similarity Index. In general, the similarity score is used as measure of how much a manuscript is similar to previously published text.
What do journals see when they use this system?
The iThenticate Demo can give you an idea of what a user (e.g., journal editor) sees when they use the actual system. However, most authors' experience with this service will be via the iThenticate Similarity
Report that a journal may send to them if a concerning amount of overlap is identified.
How to read the iThenticate Report
The basic output of the iThenticate system that a user sees is shown below.
The overall Similarity Index for the entire manuscript is reported in the upper right. The Match Overview shows the user which database sources to investigate for possible plagiarized text and reports the similarity index and the number of similar words for those specific sources.
Next to the similarity score/index, a few of the inclusion and exclusion criteria selected by the user are reported (e.g., whether to include quoted text or the bibliography).
The report also highlights the specific sources by number and color in the main text (left side) so that a user quickly sees which areas of text may be problematic.
An example of the PDF report that is often sent to authors if plagiarism is a concern is shown below, and you can also download the example report here.
In the PDF report, the match overview is usually found after the end of the highlighted paper. There is less dynamic functionality in the PDF, but some of the database sources at the end of the PDF will take you to the relevant paper if you click on it. Alternatively, the color coding in the main text can be used to see which sentences match the numbered sources.
What level of overall or specific similarity is allowed?
The allowable level of similarity varies by journal and discipline. Additionally, sometimes a higher overall similarity score is less problematic than, for example, a paper with a lower similarity score but that also has one paragraph that that has been directly copied from another source. The Springer Nature iThenticate guide for journal editors gives a good overview of how journals approach using this report.
In general, journals are looking for the following:
- A high overall similarity index score
- Verbatim copies of strings of text at the sentence, paragraph, or section level. An author may be given the chance to revise one such instance, but multiple instances of such text across sections is considered extremely suspicious, often leading to immediate rejection (desk rejection).
- A large relative percentage of similarity to a single paper (e.g., if the overall similarity score is only 5%, but the match for one source is 4%, a journal editor will likely still investigate closely).
Similarity scores from 1-5% for any one paper may be considered acceptable (assuming no outright copying/verbatim text). However, if the entire paper is a string of these similarities, it is a reason for concern. This will often be observed in the introduction, methods, and, sometimes, discussion sections.
What text should be revised?
Please see our article on identifying and avoiding plagiarism, as well as our guide on what text to look for after receiving an iThenticate report.
What if the source that was plagiarized is your own work?
Using your own previously published text in a subsequent manuscript can be considered self-plagiarism. Please see our article on how to identify and avoid this type of plagiarism. The Committee on Publication Ethics (COPE) also has an excellent guideline article on this topic, which they refer to as “text recycling.”
In general, it is important to rephrase text even if it is from your own previous papers. The methods section is often the culprit for self-plagiarism. A simple method for avoiding copy and pasting methods is to say that you used a method that has been previously published and give the reference for it. You can then describe any differences between that previously published method and the specific
one used in the current paper.
Understanding the iThenticate report and determining how to proceed after you receive one can be difficult. We here at AJE hope that you have found this guide helpful in understanding how to read the report. Please do not hesitate to contact us if you have any questions about iThenticate or ethical publishing decisions.