top of page

THE MONITOR

  • Ananya Patil Rao

Honor Court speaks about AI, bias, jurisprudence, and more

Ananya Patil Rao: What factors does the Honor Court look for when selecting new members? How are those members subsequently trained?


Honor Court: The Honor Court solicits applications for new student members whenever a vacancy on the court appears. Students usually serve a single two-year term, and with seven student members, that means that there are usually 1-3 openings each semester. The process for reviewing applications starts with the Honor Court continuing members, including both student and faculty members, reading student applications and discussing them. The primary factors that the Court has looked for when selecting students are constructing a court that has a diversity of perspectives and adding members who they believe will be insightful in examining evidence at Honor Court hearings. For the former consideration, the only aspect the Honor Court constitution requires is that students represent the different class years, but the court attempts to consider all dimensions of student perspectives. They primarily gauge this by reading applicants’ answers to the first two questions of the application. (“Why are you interested in joining the Honor Court?” and “What unique attributes or background experiences would you bring to the Court?”) The latter consideration is primarily addressed in the third question of the application where the applicants are asked to respond to a hypothetical Honor Court case. After the court members have decided on who to put forward, the names of the students are sent to the Student Government Association to go through their confirmation process. 


Faculty members on the Honor Court are identified by a faculty committee (Academic Council) or nominated at a faculty meeting and are then elected by a vote of the faculty. Recently, Academic Council has developed a portal where faculty may indicate their interests in serving on a variety of committees, including the Honor Court. Academic Council uses this information to help identify faculty to put forward for election. Faculty typically serve three-year terms and with three faculty members, there is typically one new member each year. 


Both student and faculty members are trained through training sessions held each semester, typically shortly after the newest members are added. These trainings, which are organized by the Associate Dean of Students for Academics and the Student Honor Court Chair, focus on helping members understand the process that Honor Court cases are adjudicated by and their responsibilities in that process. The training often also includes discussions of cases that were difficult for past Courts to make decisions on or hypothetical scenarios that the Court may expect to face in the future. 


APR: The Honor Court, for student privacy reasons, generally does not release the details of its decisions. With significant annual turnover, how does the Honor Court establish precedent and its jurisprudence? What avenues does the community have to compare decisions for consistency?


HC: The Court releases its community notices from each semester. They can be found here, and are released after all cases from a semester, including any appeals, have been completed. While they are intentionally vague to protect the privacy of the participants, they do describe each allegation generally and the outcome of each case so that the community can observe the actions of the Court. 


At Honor Court hearings, the Associate Dean of Students for Academics is tasked with collecting and presenting recent cases that are similar to the allegation and describing them to the Court. Typically, this is utilized mostly when determining an appropriate sanction after a student has already been found responsible, but past interpretations of the Honor Code can also play a role in determining whether a student is responsible or not. 


APR: How does the Honor Court ensure that its judgments are free of racial bias, given that a large proportion of its members are white?


HC: All individuals possess inherent biases, some of which they may be aware of and some of which they may not recognize consciously. Because of this and historical factors, institutional structures can be susceptible to bias, including, but not limited to, biases based on race. Research suggests that one of the most effective ways to minimize the effects of unconscious bias is to discuss and acknowledge that these biases exist. If one considers and thinks intentionally about their own biases, it is easier to reduce their effects. This conversation is a typical part of our semesterly training with new and continuing members. In addition, adhering carefully to established processes and protocol can be a way to limit the effect of bias. In the Honor Court context this means that when members are considering whether a student may or may not be responsible for a violation, they must start with the presumption of innocence and only find a student responsible if the evidence is “clear and convincing.” Limiting the Court’s deliberations to an examination of evidence makes it more difficult for biases to have effect. Note the court does not make decisions on whether a student is a good or moral person. Rather, the Court limits itself to whether the evidence of a student’s actions are clearly and convincingly a violation of the stated Honor Code.


An additional structural bulwark against the effect of bias is the student’s right to appeal any Honor Court decision to the Appeals Board, a group that contains both student and faculty members. The Appeals Board entertains appeals based on four possible reasons, and one of those reasons is “prejudicial bias on the part of the hearing body.”  

 

APR: How did the Honor Court develop its AI-use guidelines, and what additional training (if any), did its members have to undergo?


Note that the Honor Court has not developed any guidelines on the “use” of generative AI. How students utilize generative AI tools is completely determined by the professor of each class a student is in. In charging violations of the Honor Court stemming from AI usage, the court references section II.5 of the Honor Code that describes that it is a violation to collaborate “without acknowledgment and explicit permission of the instructor”. This means that a professor may give permission to use these tools to whatever extent they choose, and if the student acknowledges their use, it would not be a violation of the Honor Code. If a student acknowledges their use of these tools, the Court does not view it as a violation, although they accept that an instructor has full discretion for how they  may choose to grade an assignment that was produced using generative AI tools.


The Court has not received any specific training around generative AI other than discussions amongst the Court about relevant cases that have been brought. As one might imagine, this has been a significant topic at trainings and other Honor Court meetings.  

 

APR: What evidence does the Honor Court currently require in order to determine whether its AI-use policy has been violated?


HC: There are no specific types of evidence that the Honor Court requires to find students responsible for violating the Honor Code. As mentioned above, the Honor Court does not have an “AI-use” policy, but rather examines evidence as to whether a student received “explicit permission” to use any tools and whether or not the student “acknowledged” their use. 


APR: Many so-called ‘AI detection’ websites have produced a high frequency of errors. In the absence of software, what processes will the Honor Court use to determine whether a violation of AI rules has actually occurred?


HC: We would like to reiterate that the Honor Court does not recognize AI detection tools as providing “clear and convincing” evidence. The Honor Court is unaware of any tool that has the necessary accuracy to meet this standard. 


Unfortunately, there is no single process that the Court can use to determine whether a student has violated the Honor Code. Note that this is not unique to possible uses of AI tools, and just like in cases of plagiarism, cheating, etc. where a student denies violating the Honor Code, it can be difficult to determine the appropriate decision, and if a student is found responsible, to determine the appropriate sanction. After hearing the evidence of a case, the Honor Court deliberates solely on the question of whether the student is responsible first, and only after that has been determined, do they consider what the appropriate sanction should be. These deliberations are usually quite lengthy (several hours per case) because the Court needs to consider the evidence carefully and the Court is asked to think about the case from all perspectives. As mentioned before, the evidentiary standard is “clear and convincing”, which is a higher standard than “more likely than not”, but a lower standard than “beyond a reasonable doubt”. A two-thirds vote of the Court is required to find a student responsible for a violation, and sanctions are determined by a majority vote.   



APR: Of the students reported to the Honor Court, what proportion have been found guilty of violations? Data for the last few years would be helpful.


HC: When the Honor Court receives an allegation, there are three possibilities: the case can be dismissed, the student can be charged and go through an administrative hearing, or the student can be charged and go through a full Honor Court hearing. For most allegations that are brought to the Honor Court, students are charged and proceed through an administrative hearing. To be granted an administrative hearing, students must admit responsibility to the violation. Therefore, all students who go through the administrative hearing are found responsible. One can gather this information from the community notices that were referenced above. We will be releasing the Fall ‘23 notices soon, but can preliminarily report that there were no full Honor Court hearings in that semester. In the Spring of ‘23, there were four cases that led to Honor Court hearings involving five students. Two of the students were found not-responsible for any violations; one of the students was found responsible of violating section II.5; one of the students was found responsible for violating section II.5 but not-responsible for violating section II.1; and the final student was found not-responsible for violating section II.5 but responsible for violating section II.1. 


Sophia Katz, Class of 2025

Emily K. Harrison, Assistant Professor of Theatre

Claire Mouflard, Associate Professor of French and Francophone Studies 

Adam Van Wynsberghe, Associate Professor of Chemistry and Associate Dean of Students for Academics

Keelah Williams, Associate Professor of Psychology

Annika Benn, Class of 2025

Justin Kang, Class of 2027

Meredith Brubaker, Class of 2026

Arjun Mehra, Class of 2026

Enzo Battaglia, Honor Court Chair, Class of 2026


130 views0 comments

Commentaires


Post: Blog2_Post
bottom of page