Updated 11/04/2016
Let us say it at the outset: similarity detection software is indispensable for avoiding the large-scale fraud that occurs in environments where there are no checks. By no means do they claim to eradicate the problem by “treating” each knowledge delinquent at large in academia, whether that be a student or a renowned researcher.
If your institution does not have a subscription to a similarity detection program, you can look for one here and request a free trial.
The procedure is generally as follows:
- You sign in using your e-mail address.
- You then receive an access code for the document analysis service.
- You choose one or more documents to analyze and/or compare.
- The document is loaded into your workspace.
- The analysis uses the documents that you enter, but also all the referenced documents on the Web.
- You receive comprehensive and precise results, such as:
– the percentage of similarity of the document analyzed
– all the passages that are similar to those of other documents (they will appear in a different color)
– precise references for the source documents where the similarities were found
– all similar sources, sorted by probability - It is then up to you to determine what actually constitutes plagiarism of protected works.
Are there any legal issues associated with the use of these programs?
While there is no legal problem in Europe with the use of similarity detection programs in a university or school context, university heads in Canada and the US often fear using them and refer to their Constitutions.
Turnitin addresses Canadians’ concerns about legality.
Another response can be found in the recommendations of the CREPUQ working group.
CREPUQ – Rapport du Groupe de travail sur le plagiat électronique (Report of the working group on electronic plagiarism) December 2011, Québec.
How should I choose a similarity detection program?
The following advice was prepared with assistance from Frédéric AGNES – frederic@sixdegres.fr
1. Is the program easy to use?
Will you be the one using program? If so, you should find it appealing and ergonomic. It must meet your needs and the information you need must be easy to find.
It is often possible to download a trial version or to do a free test. Don’t skip this step! It will help you form your own opinion on the program
2. How does the program analyze a document?
Some programs analyze documents directly on your computer. Others require that the document be sent to a remote server, be uploaded to a virtual site, or be sent by e-mail.
Software installed on your computer:
The program will remotely search databases on the Internet. If your school’s network has a firewall, the program might not be able to send its queries.
Software that can be used online:
Users have personal accounts where they upload the documents to be analyzed. Some programs even let students upload documents directly. It is usually through the same account that you can check the results of the analysis.
3. How much time does it take to get the results?
There is variety here as well. Some programs are very quick and need just a few minutes; others are much slower and need a few hours or even a few days.
Software installed on your computer should in theory take less time, but your computer will have to be turned on for the duration of the analysis.
Online programs can continue the analysis of your documents even when your computer is turned off; everything happens remotely on the program’s server. Then it’s just a question of seeing how much time the programs take to get you your results.
4. What type of results will I get?
• Percentage of similarity
The first thing you will want to know is the percentage of similarity between the document analyzed and the sources found.
But beware: do not confuse similarity and plagiarism: quotations and some very common expressions will be considered to be similar but will not constitute plagiarism.
• Similar passages
The percentage of similarity is calculated according to the amount of authentic text as compared with the amount of similar text found. Most programs show you the similar passages so that you can decide for yourself whether or not it is a case of plagiarism or a quotation.
• Sources
Secondly, the programs tell you the sources where each similar passage was found. Some programs even rank the sources according to their frequency.
5. What sources do the programs find?
Here as well, the programs take different approaches.
Some build their own databases by indexing websites, documents found on the Internet and documents submitted by users.
Other programs are meta-search engines: they don’t build their own database of documents but use online search engines and bring together the results.
Yet other programs combine these two approaches.
In all cases, content that is freely available on the Internet will be found. In order to access texts on sites that require a subscription or the payment of a fee, publishers of anti-plagiarism software must establish partnerships with the distributors of those texts.
6. Do the programs find translated or reformulated sources?
Most programs find extracts that are identical. They will therefore not find passages that have been rewritten or translated. In both cases, it is up to you to determine if there could still be a case of plagiarism…
Some programs do offer a semantic verification service that substitutes words or phrases with synonyms or even translations. In practice, however, a “human” translation is rarely similar to the possible translations suggested by the programs.
7. Can I fine-tune the results?
You might read the results of the analysis and notice that an extract isn’t due to plagiarism but comes from a quotation. You also find that a large number of extracts come from sources that you yourself had recommended to your students. What can you do?
Depending on the approach used, some programs let you adapt and adjust the results of the analysis. You can, for example, ignore a source or a piece of text in order to recalculate the percentage of similarity.
In order to do that, you need to be using an interface that allows you to modify the results. If the results are presented to you in a static document, you will of course not be able to modify them.
8. Where am I going to use the program?
- If you often use different computers (for example, at work and at home), you should rule out programs that are installed on a single computer.
- If you use an e-mail program installed on one computer to access your e-mails (e.g. Outlook, Eudora, Thunderbird) and you want to be able to access the report on several computers, you should probably not choose a program that sends results by e-mail.
- Finally, if want to consult your reports on computers that are not connected to the Internet, you should not choose “virtual office” type programs.
9. How will I collect the documents to be analyzed?
Perhaps you have used a program to analyze one document and the results seem conclusive.
Now let us consider what happens when You have to collect a hundred documents to have them analyzed. There are several possibilities available, and all the possibilities are not offered by all the programs.
Which option is the best suited to your needs?
- Collecting documents (on disks, by e-mail, through the university e-portal, etc.)
- Or collecting documents and then uploading them onto your account
- Or asking your students to submit their texts on the program’s website. The documents will then be directly available in your account.
10. Do all the programs have a deterrent effect?
The element of the plagiarism detection programs with the greatest deterrent effect seems to be having the students send or upload their papers themselves. They then become direct participants in the verification process and are immediately confronted with the reality that their work is being checked.
For this reason, the programs that respond immediately, saying “Your document has been received” or “Your document has been successfully uploaded” are to be preferred for their increased deterrent effect.
What are the limits of similarity detection software?
While search engines do help identify cases of plagiarism, those cases will involve plagiarists who are either naive or who are counting on the naiveté of their professors. And the use of search engines has significant shortcomings:
-
Search engines do not have access to the content of sites selling ready-made papers, such as “Pickdoc,” which then became “Oboulo” and is now DOCS.school.
-
They also do not have access to password-protected pages. And it just takes a few seconds to create an free and protected interactive website (a “private group”) on MSN or Yahoo!, and it can then be used as a platform for selling or exchanging papers.
-
In addition, it is enough, for example, to replace a space between two words with two spaces or to reverse the order of some words in order for the search engine to not recognize the quotation. On bien sûr d’inverser les mots.
-
There are now even paraphrasing websites that can be used to avoid detection.
These shortcomings can be exploited in seconds by anyone, even someone without any special IT knowledge. In addition, copying and pasting a completed text is not the only means of plagiarizing.