This is a widely shared but total misunderstanding. Software is used to detect similarities in a student's submission, but it is academics who must decide if it is plagiarism. In dealing with what is a major problem (students presenting other's work as their own) there are two separate issues.
1. It is not plagiarism if you put something in quotes and correctly attribute it. However, it is not original work if you simply include lots of quoted material within what is supposed to be your work. Thus institutions increasingly require anything at the level of Master's or PhD to be a high percentage of the student's own words, with quotations used only where necessary. It is expected that researchers will provide a critical summary of the paper, not just "reference A said........". My institution's standard is no more than 5% as quoted.
2. It is practically impossible to write a comprehensible thesis in entirely original words. Thus a computer program will find matches for similar and even sometimes identical sentences in works which the writer has never read. There are settings which academics can use, but generally it is rare to see less than 5%. More than 20% is suspicious, but can occur innocently if, for example, there are papers included as an appendix. However, in my experience a student can be expelled for plagiarism with much less detected by the software. This typically occurs when a student copies the structure of a publication but changes words (e.g. large into big) in order to disguise the fact. This is of course an intention to mislead, so deserves maximum punishment. In another case a colleague was External Examiner for a PhD and recognized substantial portions of her own PhD!
Plagiarism of whatever percentage is still plagiarism and should not be allowed. For an academic/scholarly work such as thesis or dissertation, universities and education commissions may allow a certain percentage (say, less than 20%) of "similarity" to previous works but there should be 0% plagiarism. To avoid plagiarism, proper citation of previous works or ideas must be done.
Totally agree with Romer Castillo . Plagiarism even if 1% is still plagiarism and totally immoral and illegal. However, we shouldn't confuse plagiarism with similarity index. The Similarity Index does not indicate plagiarism, rather the amount of text that has been found in other published material or student papers. If the author has referenced the material then that is not plagiarism. It may not be high-quality academic writing but it shouldn't be treated as plagiarism.
Here is a reason that you may not expect. The software detects plagiarism by word matching to a massive data base. Easy. But It can NOT tell whether what it thinks is plagiarised in the work it is checking, has been correctly referenced or not. For this reason an 'allowance' of 20% is made and this means every piece of work is identified as containing plagiarism and it is up to the instructor to determine the correctness of the citations and referencing. This is also a covert means for the maker of the software to generate amazing statistics that show how valuable their software is. I refuse to use it. Why? Because I actually read every piece of submitted work and make my own judgements
If my earlier suggestion is valid, then it is expected a paper submitted by a student will be about 1/5 based in the literature nd 4/5 their original work explaining what they cite. I know it sounds silly, but to me it seems just a 'best guess' that up to 20% of a piece of work can be cited and 80% original. I do not think it is a hard and fast rule, just some mrleting person's 'idea' ... dreadful, isn't it?
Phillip Ebrall agreed with your both answers, 20% similarity and 80% original work of author. I think proper citations should be done to avoid plagiarism.
Here is another answer that is different to what you may expect from an experienced (30+y) academic: I am not concerned with the format of citations used by learners. Why? Every journal has its own style, so who am I to insist on any one style which I might happen to like? It is equally immature of a university to insist on one 'style' for the university. Why? Every discipline has its own literature base of journals, and even the journals in one discipline use different styles. Two things are important: (i) the learner demonstrates they know how to search, find and document their evidence, and (ii) they look widely and really 'search' the literature instead of platitudinously throwing up a handful of v'standard' references.
This is a widely shared but total misunderstanding. Software is used to detect similarities in a student's submission, but it is academics who must decide if it is plagiarism. In dealing with what is a major problem (students presenting other's work as their own) there are two separate issues.
1. It is not plagiarism if you put something in quotes and correctly attribute it. However, it is not original work if you simply include lots of quoted material within what is supposed to be your work. Thus institutions increasingly require anything at the level of Master's or PhD to be a high percentage of the student's own words, with quotations used only where necessary. It is expected that researchers will provide a critical summary of the paper, not just "reference A said........". My institution's standard is no more than 5% as quoted.
2. It is practically impossible to write a comprehensible thesis in entirely original words. Thus a computer program will find matches for similar and even sometimes identical sentences in works which the writer has never read. There are settings which academics can use, but generally it is rare to see less than 5%. More than 20% is suspicious, but can occur innocently if, for example, there are papers included as an appendix. However, in my experience a student can be expelled for plagiarism with much less detected by the software. This typically occurs when a student copies the structure of a publication but changes words (e.g. large into big) in order to disguise the fact. This is of course an intention to mislead, so deserves maximum punishment. In another case a colleague was External Examiner for a PhD and recognized substantial portions of her own PhD!
I have cited and referenced all my research papers with direct links to original sources while writing papers with my original thoughts and still ended up with less than 10 percent plagiarism rating from similar software. The reality is that these software, institutional policies and thresholds are not all correct or perfect; however, some standard must be set to discourage students from attempting to plagiarize and steal the work of others.
A very good question and finding, if you allow me, the 20% in my opinion is equal to 1/5 of similarity in the text submitted with other texts already published, which is very logical and very digestible especially for those who publish for the first time, which they need to draw from discussions from other studies.
I would add a question to yours: why not change the percentage of plagiarism for experts and make it 0% for those with more than 10 publications for example? That way it will be very fair and logical.
Again, I think it is about maintaining integrity and holding all researchers and academics accountable for their published works, regardless of how many they have previously published. Even though, most of us believe in truth, intergrity and honor among us, there are still some who do not.
This 20% has been standardized by several academic instituions, scientific centers, journals and etc. More lesser than 10% would be much welcomed. To me as long as the used work or the written sentences are cited with the right reference and source then that would be acceptable even if it is more than 20%. This is just my opinion. Overall, this is very excellent question and well shared.
Please consider that plagiarism is not acceptable at all.
The percentage you have mentioned relates to the similarity in words or some other things except the core idea. Besides, it is noteworthy that each journal has its own criteria to determine the similarity index which can be different from one journal to another.
Abdel Nasser Umran says plagiarism of 20% has been standardized by several academic instituions, scientific centers, journals and etc. More lesser than 10% would be much welcomed. To me as long as the used work or the written sentences are cited with the right reference and source then that would be acceptable even if it is more than 20%. This is just my opinion. Overall, this is very excellent question and well shared.
On the other hand, Martin John Pitt talks about the other side of issue which is a widely shared but totally misunderstood situation. Software is used to detect similarities in a student's submission, therefore it only detects similar sentences and words. But the major problem is not using the same words. It is the problem that a student presents other's work as their own. These are two separate issues. Therefore, it is not plagiarism if you put same words in sentences and for that you do not have to quote and attribute. That is pretty stupid. What should be quoted is that the main idea the main claim in the discussion part. We are not dealing with if it is an original work or not. That is the responsibility of supervisor to decide. if you simply include lots of quoted material within what is supposed to be your work. It comes to a point that one hould quote for every sentence he/she writes. This pretty ugly situation. Therefore, only the discussion part should be quoted just because of it is the pure creation of the owner of the thesis. Otherwise, Just in the introduction, the writer starts with quotation of every stupid words and sentences he/she writes. This is now the situation in Turkey!
Thus institutions increasingly require anything at the level of Master's or PhD to be a high percentage of the student's own words, with quotations used only where necessary. But words not own by the students, anybody can use them. I now use the sentences ofMartin John Pitt but we do not go to the same aim. My aim is different therefore I do not have to quote that Pitt did this before. That is none sense. But in conclusion I agree with Pitt and now quote his words; "It is expected that researchers will provide a critical summary of the paper, not just "reference A said........". My institution's standard is no more than 5% as quoted."
As a result, it is practically impossible to write a comprehensible thesis in entirely original words. Thus a computer program will find matches for similar and even sometimes identical sentences in works which is not important that the writer has read it or not. Just one stupid sentence is now accepted as plagiarism!
Therefore Romero Castillo's words that plagiarism of whatever percentage is still plagiarism and should not be allowed is nonesense. Since we are not dealing with "similarity" to previous works but we are dealing with words and sentences. Therefore computer programs' matches nothing but stupidity. If two thesis are similar that is something different as Pitt says. So be carefull about it.
Plagiarism is plagiarism whatever is the percentage of similarities. Did I now make plagiarism from Safia? Brains stick to percentage can not solve this question?