I've studied at an online university. There were quizzes every week that were not graded; one or two graded ones during the course; one final graded quiz at the end of the course. 99% of all the questions were multiple-choice ones. Some questions from non-graded quizzes would appear in the graded ones. Then the questions from the graded quizzes during the course would appear on the final one. After taking the quiz, you could see the correct answer for each question. As I went through the course, I would add every single quiz question I encountered, the answers, and the correct answer into the database. Sometimes I would google questions I saw on the quiz and would find the list of the questions somebody else had posted. I would then verify the correctness of the answers against the course material and add them to my database as well.
I wrote a JavaScript that would get injected into the page with a quiz using the GreaseMonkey extension. It would go over the questions on the page, do a fuzzy search against my database that I compiled, and put the top 5 best matches next to each question. The library had its limitations, but it would take care of things like words being in a different order or a couple of words missing. I improved it by adding additional transformations to the question. Things like removing any non-letter characters but being careful to handle things like dashes and apostrophes in some cases; removal of the stopwords; stemming.
That would usually take care of about half of the questions. Sometimes, more than 90%. I would then spend the rest of the quiz time manually answering the remaining questions. Since this course was an open textbook, I would have one giant text file that would contain every single text that appeared during the course. It was allowed to look into any course material while taking the quiz.
The end result of this is that I would get 98+% of the answers correct in almost every single course I took.
We also had written assignments which I always honestly completed while carefully following the citation rules. No clever tricks here.
I graduated with 10% in my university and was on the honors list. I never thought of what I did as cheating because I never collaborated with others. I never disclosed what I did and how I did it. I think that people who did those quizzes did them in a way that allowed me to do what I did.
I wrote a JavaScript that would get injected into the page with a quiz using the GreaseMonkey extension. It would go over the questions on the page, do a fuzzy search against my database that I compiled, and put the top 5 best matches next to each question. The library had its limitations, but it would take care of things like words being in a different order or a couple of words missing. I improved it by adding additional transformations to the question. Things like removing any non-letter characters but being careful to handle things like dashes and apostrophes in some cases; removal of the stopwords; stemming.
That would usually take care of about half of the questions. Sometimes, more than 90%. I would then spend the rest of the quiz time manually answering the remaining questions. Since this course was an open textbook, I would have one giant text file that would contain every single text that appeared during the course. It was allowed to look into any course material while taking the quiz. The end result of this is that I would get 98+% of the answers correct in almost every single course I took.
We also had written assignments which I always honestly completed while carefully following the citation rules. No clever tricks here.
I graduated with 10% in my university and was on the honors list. I never thought of what I did as cheating because I never collaborated with others. I never disclosed what I did and how I did it. I think that people who did those quizzes did them in a way that allowed me to do what I did.