IYCC for Geeks "Geeks making geeks smarter."

The argument is frequently made that assigning a student a failing grade and/or expelling the student is a sufficient deterrent to plagiarism. Game theory suggests that such a policy is no deterrent at all and that such policies make plagiarizing a winning strategy for ill prepared students.

The following analysis makes the assumption that the student cannot pass the course without plagiarizing. I believe the assumption to be reasonable because plagiarists spend at least as much effort plagiarizing as it would take to do legitimate work. In other words, plagiarists aren't strictly speaking lazy. In 5 of the 6 cases in my course, the plagiarists could have done less work and gotten a better grade had they simply done the work according to the directions in the syllabus. (For this analysis, I'll ignore intellectual laziness. Regurgitation of conventional wisdom is a well-tested path to tenure and subsequent promotion to faculty administration positions.)

The second assumption is that the course is required to graduate. Few students take courses unless they must, and those that do enjoy doing the work and score well.

The third assumption is that the actual incidence of penalties is uncertain, for two reasons. First, detection is uncertain. No search engine or software program can be 100% effective. In addition, a variety of gambits are available for evading detection, such as: rewording or rearranging the source, finding obscure sources, using fee-based services, or obtaining private assistance from lovers, sorority sisters, friends, etc. Second, the secretive nature of University proceedings and Honor Committees introduces additional variability in the outcomes. If the process is not transparent to the public, the attractive, glib, manipulative, connected, or rich student can influence the outcome and avoid penalties.

Assume two students who, as stated previously, cannot pass the course without plagiarizing. The first student nonetheless attempts to do the work without plagiarizing. The second student plagiarizes.

Which has the best strategy? Under these conditions, the plagiarist wins.

The first student turns in the work, receives an "F", and fails to graduate. Thus, we score the honest student with a zero return.

On the other hand, the plagiarist turns in a plagiarized work. If caught, the plagiarist presumptively receives an "F" and is expelled, subject to the plagiarist's ability to mitigate the penalties discussed below. Because the course was necessary to graduate anyway, expulsion no different than simply failing to graduate. Thus, the plagiarist is in no worse position than the first student.

However, because detection is uncertain, the plagiarist has the possibility of not only passing but receiving a good grade. Plagiarists usually can't tell a good essay from a bad one, so the expected outcomes would be distributed among the various possible letter grades. (I don't have any empirical evidence what the distribution would be, but my sense is that it would approximate a discrete uniform distribution.)

The plagiarist's expected outcome is:

• the summation, for each respective letter grade (A, B, C, D, or F), of
  • the probability the plagiarized work receives the respective letter grade,
  • multiplied by the value of the respective letter grade
• multiplied by the probability of escaping detection.

Thus, as long as the probability of escaping detection is non-zero, the plagiarist's expected outcome exceeds the first student's expected outcome.

The analysis can be further extended by factoring in the mitigating effects of the possibility that the secrecy of proceedings enables the plagiarist to attenuate or eliminate the effects of getting caught. Essentially, a secret process that allows for input by the student gives the student a "second bite at the apple" by enabling the student to affect the outcome. The formulaic expression of the mitigation is left as an exercise for tenured faculty who don't actually grade students work.

Thus, plagiarizing really does pay if the only penalties are altering the grade and/or expulsion. If the adjudication process is secret, the plagiarist gets an added bonus because the plagiarist has the opportunity to mitigate consequences.