ProctorFree Research – Post 3

My research has officially come to an end.

Unfortunately, my second test plan will go unused. After the midterm exam in Professor Schreiber’s class suffered from Proctor Free related technical difficulties, the final exam was conducted without the service. Since we lacked support from ProctorFree, there was no other way that we could access the system.

This is especially tragic because the majority of my research was geared towards that second test, and it was far more technically advanced and significant to the general conclusion.

Despite that, I still gained a wealth of information from the project.

First, I can make reluctant judgements on the effectiveness of ProctorFree’s service from the results of my first test. Of course, I must state that I do not have enough data to make a conclusive statement about the system. Had Dr. Kemper’s data been returned, I would have been able to be slightly more definitive in my conclusion, but the status of those results is still unknown. That being said, my results were not entirely insignificant. While I only did one “test run”, in that one protocol I tested ProctorFree’s facial monitoring system six times: using a cheat sheet, student collaboration, a cell phone, a textbook, copying down test questions and taking a “bathroom break”.

Since not a single cheating attempt was caught, it is fair to say that ProctorFree’s facial recognition software leaves something to be desired.  Once again, due to the issues with our data, I cannot conclusively say that ProctorFree’s student monitoring software is completely faulty. However, six missed positives in a row is not a good sign, and would cause me to hypothesize that if more data were to be collected, it would show significant gaps in the system.

I also gained important personal experience from this endeavor.

Most glaringly, I learned the perils of doing corporate research whilst not part of said corporation. As in this case, corporations make decisions that are difficult to predict from an individual’s perspective.  As a result, issues such as those in this research are unavoidable and unfixable due to potential lack of access to the research materials.

On perhaps a more scholarly level, I gained experience with the general process of software testing, an extremely important skill for a computer science major.  Through my research into software testing, I learned different protocol formats and how to determine what should be tested in a piece of software. Going forward for future research projects, I expect that this skill will be the most valuable product of this research, as I plan to continue in the field of computer research.

In closing, this research shows the importance of student integrity and strong school honor codes. Through my research into cheating in higher education, it became clear that cheating is simply a part of life. A professor could never hope to catch every cheater in the act, and as we have discovered, a computer couldn’t either.  So how do we keep the integrity of college degrees? How do we ensure that when a person earns a degree that they actually earned it? The answer is to stop cheating before it occurs. Creating a strong honor code and accepting students only with high levels of integrity is simply the best way to prevent cheating in a college. With the proliferation of technology, it is becoming harder and harder to catch a cheating student, especially since cheating prevention technology lags so far behind. The only way for an institution to stay well respected in a world where getting ahead is becoming more and more imperative is simply to create a student body that is dedicated to hard work and fairness, not a computer program to force them into this mold.



  1. dvpontari says:

    Very interesting research idea. This project kind of reminds me of something Frank Abagnale from “catch me if you can” would do. I wonder what type of quality assurance ProctorFree has in house. I am sure they would be interested in learning about your findings. From a business perspective, I am not sure how important it is for Proctor Free to spend the time and money installing additional complex safety measures for the small percentage of students who can program. I equate the added security measures to installing shatter-proof plexiglass windows in cars. If companies were to put in sturdier glass in cars, we could probably have fewer break-ins. However, there is only a small percentage of people who break into cars through windows and for car companies, the value of shatter proof windows to the customers is not worth the cost in production. Economically, I would assume the same holds true for added security measures in proctorfree. The service is good enough and has good enough safety features for professors. Still, I would be curious to see if proctorfree offers another service that has even better security features to accommodate computer science classes.

  2. iechevarria says:

    Hi Philip, I really liked your project. I think it’s awesome that you actually tested a piece of software that is used by large institutions. I’m surprised that you were able to circumvent the cheating detection with every attempt, and while there were only six tests, they were very rudimentary and they probably shouldn’t have succeeded so easily. Do you think your research will have any effect on the adoption of ProctorFree as a solution for monitoring students while they take tests online?

    Even your brief tests seem to suggest that ProctorFree doesn’t work, and I think it’s awesome that you discovered this.