The Academic Fraud of Non-Proctored Exams

Did I catch your attention -- good! Now, I don't mean to imply that all non-proctored exams are bad; however, my experience tells that that some non-proctored exam practices enable academic fraud. Let me give an example of what I mean:

I've had classes at both Saint Petersburg College and Florida State University (two @ SPC and one @ FSU) where the student was given an online exam where all they needed to do was to login and complete a mostly if not all multiple choice exam in a fixed amount of time. In each case the student was required to complete the exam without using their books, notes or assistance of any kind. Because the exams were multiple choice and were taken directly from the notes or the text, my trust in humanity tells me that a lot of cheating was going on. Does anyone disagree? As for myself, I found the lack of effort to develop a respectable testing regimen to be utterly depressing.

Do I think all non-proctored exams are easily beaten by frauds? No, let me give you an example. Last semester I had a graduate course in Artificial Intelligence. Both the mid-term and final exams were non-proctored, take-home tests. From the moment they received the test, each student had a 24 hour time limit to return the test to the professor. The test was administered to both on- and off-campus students exactly the same. Each student could use any resource they wanted (textbook, notes, Internet, etc.) These tests were long and many of the problems were extremely difficult requiring the creative synthesis of knowledge of course material and general knowledge expected of a graduate student. I worked for 12 hours on the mid-term and 14 hours on the final. In each case I sweated my grade. Now *these* are non-proctored exams I can respect!

In addition to take-home exams similar to the ones described above, I can mostly respect tests in the form of term papers that require the student to analyze and in the process demonstrate a command of the subject beyond a simple regurgitation of the text and lecture notes. Like my take home exams, a lot of learning goes on while taking the test. Term papers aren't perfect but they are a far cry from multiple-cheat tests. All-in-all, it’s my opinion that most non-proctored testing regimes as simply and pedagogically millish.

 

I agree with you. But i would'nt call this "Academic Fraud". Last semester most of my exams at TAMUC were take-home case study based. They were not easy, it took me around 15 hrs of research and study to complete each one of them. This actually make the students to study "in-detail" and do research on the subject matter.

 

I don't think all non-proctored exams enable fraud, but some do. Tests that are designed to be take-home exams (like the one you describe) are perfectly translated to an online environment. However, exams whose design mirrors exams that should only be given under the watchful eye of an instructor or proctor but are instead given online, seem very suspect.

 

My last exam in the Liberty University MBA program was 50 multiple-choice questions. After logging in to the exam online, we had one hour to complete the exam in one sitting and were penalized for each minute over the one-hour time limit. The exam was open book but the one-hour time frame did not leave a lot time for searching through your textbook or notes. I believe this is a good approach.

 

The exams from Penn Foster are open-book, online exams, with the exception of the end-of-semester exam, which is open-book, but proctored.

I thought the online exams were going to be a piece of cake, but they are not. Sure, the exam is open-book, but the questions are worded in such a way that the answers are not easily found in the book. They require you to take what you have learned from the book and apply them to situations you may not have read about in your text.

I study long and hard for my exams. Our online exams are not timed. My last Business Math exam took me well over two hours to complete, and it was a 20-question, multiple-choice, open book exam!

 

I have written both proctored and non-proctored, but timed, exams and neither is easy. Multiple choice exams are not easy whether open-book or closed-book. Essay-based exams which are supposed to reflect synthesis of subject area knowledge and critical thinking need not be proctored. Mathematics exams whether open-book or closed-book rquire you to know the material before starting the exam. Good instructors can catch cheaters (those having others write their exams) based on previous performance in a course.

 

Thank you for these thoughts...

I'm writing my syllabus and exams for my first (!!!) teaching assignment. Quizzes will be online but the exam will be proctored w/ID required and will be open note but not open book. Thus the students will not need to memorize formulas but will need to be able to apply the concepts to the questions and real-life.

My mentor/boss has told me that she develops a sense of the student's "voice" in online postings so when she receives the term papers, she can usually tell if the writing is not that of the student's (either purchased term paper or plagiarism).

And a large difference between an exam grade and quizzes/paper grade can call into question the honesty of the student's work on the non-proctored portion of the assignments.

This is a good discussion.

 

Southwestern College mixes it up a bit. Almost every exam I have had is timed, every Final is timed and they also schedule a final paper to go with the exam. I would find it terribly difficult to cheat, even with a book, 30 questions in an hour to an hour and a half would be difficult if not impossible if you did not complete the readings.

As for the papers, well, anyone in a classroom has about as much if not more opportunity to cheat than those online.

 

Whenever a non-proctored, multiple choice exam is given closed-book, closed-notes, assuming that students are honorable is simply ostrich-like. Cheaters are not just people who don't do the assignments or read the material, they are also people looking for any advantage to keep their GPA high. Consider this -- if the student has read the material, then it just makes it easier to cheat because they know where to look for the answers! All non-proctored exams should be designed to be open-book/open-notes/open-Internet/open-associate etc. or else I personally can only assume that the results of the test are fraudulent.

What do you guys feel are the best types of online exams that prevent or at least minimize cheating? I've already mentioned the take-home test from hell or term papers. I personally think all multiple choice questions are lame. Of course they are easier to grade and they do test a certain kind of understanding of the material, but they do not test the creativity and synthesis of knowledge that a graduate student is expected to have. After all, it is much less creative to work backwards given a small set of answers where one (and only one) of the answers are expected to be correct.

To maintain their credibility, DL schools should design all assignments with the assumption that students will beg, borrow or steal to get the best grade they can.

 

At Capitol we have at least one proctored exam a course. Other exams are always take-home or done online via blackboard. Most of my classes are highly deliverable based, with 2 – 3 exams a class. All tests are open note, open book and you can even play back the old lectures via Centra. For those of you that do not know, Capitol has live lectures. Each class has at least 8 live 3.5 hour lectures, sometimes more.
The proctored exam makes sure no one else helps you; however, they are always a lot shorter simply because the others can take more than 20 hours to complete. It should be noted, the proctored exams have time limits. If you do not know the material well, you will mostly likely run out of time.

To give a slightly larger view of my current experience, in one of my current classes (NS680 - Perimeter Protection) I spent 18 hours on my first project (plus the write up), of which we have three. The first requires an intimate knowledge of TCPIP, vulnerability assessment tools and ethereal. My second lab requires the use of Snort and related tools. The requirements document alone is over 20 pages.

The point of my babble is it depends on the context, the format of the class and the exam. Most of my un-proctored exams require 2 - 3 paragraph responses and generally have 30 – 50 questions. Could I BS it or cheat? Sure, but it would take just as much work to do it right and learn it myself.

 

For my courses, Statistics and Writing, I think the best tests are those that measure the student's ability to synthesize information drawn from various sources, as they most closely reflect the desired course outcomes.

When the student has open sources and plenty of time to use them, the tests have validity because, rather than correlate with the course criteria, they exemplify them.

Valid, reliable multiple-choice tests are very difficult to design. The work involved is greater than the work involved in grading essays and projects.

I favor letting students use any sources they wish and giving them plenty of time to do so. Then the results tend to be of higher quality, and consequently easier to grade, as well as better measures of what students can do. Among my students, there are such differences in how much they have learned that it is easy to distinguish grade levels.

As far as plagiarism is concerned, it is usually easy to spot and document statistically. However, one school I work for will not uphold a plagiarism charge unless the original source is produced. Fortunately, I guess, my students who plagiarize are no more competent at using sources illegitimately than they are at using them properly.

 

UMUC and Park University require that all of their students have their exams proctored by an approved testing official, in an approved testing facility.

Other schools, like NCU, Capella, and Walden, simply have a final course project.

At OU, many of my professors had us complete take home exams.

And all use turnitin.

 

Good info.

Unfortunately, turnitin.com won't ordinarily spot papers purchased and/or written specifically for the student by a proxy. That is where I hope the required message board postings will help me determine the student's true "voice" and spot work that might not be their own.

 

JoAnn, while I understand your concern about the integrity of the academic process it would seem unfair to assume, that given the opportunity, students will cheat.

I am of the school of thought which prefers students who learn to think and reason rather than memorize facts. In the distance learning realm students should be expected to write papers and/or prepare projects demonstrating their mastery of the subject matter. For example, the general education courses I completed last year could not have been done in 8 weeks had the requirement been to write closed-book exams (multiple choice or essay questions). Yet, I learned a lot from these courses; primarily a basic introductory knowledge about the subject. Most of the courses required comprehensive assignments; those requiring an on-line exam could not have been passed given the 8 week terms and working full-time.

 

I understand your point, but it is really unfair to students that don't cheat to assume that nobody cheats. In fact, if the grades are based on a curve, you may end up forcing people to cheat who wouldn't otherwise cheat. When the testing method enables cheaters, then all the grades are highly suspect.

 

The BS I am in the middle of at Peru right now has no proctored exams. Does this mean it is easy, no! We have a few different assessment criteria which include the multiple guess tests, essay exams, and like you said the dreaded take home test/term paper. The essay tests that we take are more like writing a paper as they have to be 3-5 pages and have text citings. These exams can destroy you and are by no means easy even though they are open book. This is becuase the student has to synthesize a coherent argument based on the text and original thought. The multiple choice tests are easy if you do the assigned reading, although I can see that the less then prepared will cheat. This is somewhat deterred though by setting of a time limit such as 10 min for a 20 question test which doesn't leave a lot of time for page turning. In the end you are only cheating yourself anyway if you don't learn what you are supposed to. What good is having a degree if you don't have clue #1 about the info to back it up.