Charge to the Academic Integrity Working Group
The AIWG was tasked with its mission by the same university entities responsible for the Committee of 12. These include the Board on Conduct Affairs, the Undergraduate Senate, the Graduate Student Council, the Senate of the Academic Council, and the Office of the President.
Table of Contents
Background
In June, 2023, the Office of the President granted final approval to the recommendations of the Committee of 12, including revisions to the Honor Code and a mandate for the creation of the Academic Integrity Working Group (AIWG). This document serves to charge the AIWG under the shared authority of the Board on Conduct Affairs, the Undergraduate Senate, the Graduate Student Council, the Senate of the Academic Council and the Office of the President.
The Charge
The AIWG is charged to:
- Study Stanford’s current academic landscape to identify: a) the scope of the problem of academic dishonesty; b) the root causes (e.g., academic pressures, student mental health); and c) the interplay of the state of academic integrity and pedagogical practices in use. Recommend policy changes and other measures that will effectively and appropriately address these concerns.
In its April 6, 2023, report “Committee of 12 Proposals and Recommendations on Judicial Charter and Process,” the Committee of 12 spelled out a number of academic integrity topics in need of consideration beyond the proctoring of in-person exams under the section titled “Recommendations on Academic Integrity Policies” (pages 30-36), particularly the subsection titled “Proposals for further AIWG Investigation” (see report text in appendix). The AIWG will discuss which of these and other academic integrity issues it will address directly and which should be delegated to other university committees or bodies. As noted in the April 6th report, the AIWG “may reassign any of these to a standing committee at Stanford with student and faculty membership that is well-suited and agreeable to taking it up”. The university should provide resources, including access to data, to support these activities. - Carry out a multi-year study (between two to four years in duration, to be determined by the AIWG) of equitable in-person proctoring1 practices to answer student questions during exams and promote academic integrity by supervising the assessment process. The study will address concerns about proctor presence and concerns about its absence in view of Stanford’s present academic culture. It will assess the viability of in-person proctoring of exams to reduce cheating and to determine the impact on students taking exams (such as stress and academic performance).
The study will be led jointly by the AIWG and a disinterested, unbiased external consulting group with extensive experience working with student conduct, campus climate, and DEI concerns. The use of the external consulting group ensures proper and timely data collection, professional management of the data, and continuity regardless of turnover in the AIWG. The precise parameters governing the creation of this study will be determined by both the AIWG and the external consulting group. However, remote proctoring, whether by software or humans through a computer, is not under consideration.
1Here and in the approved documents from the C-12, “...proctoring” is defined as the reasonable supervision of exams by an exam administrator, and “in person” includes traveling athletes, SCPD students, and other enrolled students taking an exam off-campus.”
In the course of its work, the AIWG should study proctoring practices at other institutions, including peer institutions and others with an Honor Code, and it should take into consideration the information gathered by C-12 on other institutions.
The AIWG should incorporate student feedback collected by the C-12 process into its design and assessment processes. In particular, the AIWG should consider the best available data on bias, equity and proctoring practices. The proctoring pilot should focus on courses where cheating on in-person exams is anecdotally most suspected to be prevalent, including large enrollment courses but also some smaller courses to assess how proctoring principles apply in different course configurations. Course instructors should be consulted during the pilot design process.
The relationship between artificial intelligence and academic integrity is rapidly evolving, and a comprehensive treatment of this topic likely exceeds the scope of the AIWG’s work. However, the AIWG should include significant consideration of AI issues and how AI factors into what academic integrity means today.
The resulting data will be the basis for future policy proposals by the AIWG to the relevant authorities as identified in the Stanford Student Conduct Charter of 2023 (see section VI B, “Amendment of Charter”).
Deliverables
The AIWG will produce a final report, to be made available to the university community, that advises Stanford administrators and governance bodies on any changes that need to be implemented in relation to: 1) broad consideration of issues of academic integrity (such as those identified by the C-12 in the April 6th report subsection “Proposal for further AIWG Investigation”); and 2) Stanford’s proctoring policy within one year after the conclusion of the pilot study. If possible, an interim report on progress and data collected to date should be shared with the university community. The AIWG report will also serve to a) apprise faculty and students university-wide of the study’s results and its implications toward the goal of informing discussions of academic integrity locally (in departments, dorms, etc); and b) in conjunction with the Board on Conduct Affairs, seek out stakeholder votes as needed for the timely implementation of its data-backed proposed changes.
Membership
The AIWG will consist of four students (two graduate and two undergraduate, from diverse backgrounds including first-gen/low-income and international students, and students from diverse majors/fields of study) and four faculty/lecturers (especially from departments with the most frequent instances of in-person exam cheating concerns), along with one person each from the Office of Community Standards (OCS) and Office of General Counsel (OGC). Student appointments will be for one academic year with the option to apply for renewal. Membership in the AIWG is to be determined respectively by the ASSU Nominations Commission, the Faculty Senate Committee on Committees, and the VPSA (for OCS and OGC). It will meet with stakeholders (student groups, departmental representatives, Academic Advising, etc.) to inform its proposals. Non-voting members may include representatives from the Center for Teaching and Learning, the Office of Accessible Education and others as identified by the Working Group. VPSA will provide resources for project management.
Appendix
In this appendix, additional topics for consideration by the Academic Integrity Working Group are discussed. This represents the outcome of extensive feedback gathered by C-12 and so is important for the Academic Integrity Working Group to consider. The text below is an extract of the C-12’s report “Committee of 12 Proposals and Recommendations on Judicial Charter and Process” (pages 28-34).
Recommendations on Academic Integrity Policies
The importance of academic honesty and integrity at Stanford University. Most Stanford students are honest and work extremely hard during their time here. The value of a Stanford degree or transcript is through what it conveys about content knowledge, intellectual ability, and skills. The incentive for cheating is that a higher grade is perceived as reflecting greater content knowledge, intellectual ability, and skills.
History of Stanford University's Honor Code. A premise upon which the Honor Code rests is that students will report on academic dishonesty that comes to their attention. Indeed, the genesis of the Honor Code is a January 20, 1921 proposal for a proctorless exam system involving student reporting of infractions, in the Stanford Daily article "The Honor System" that states:
"The success of the honor system is naturally dependent on the students themselves… Responsibility for reporting infractions of the honor code must be taken by all alike, and in particular, members of the upper class must consider themselves directly responsible. It takes courage for any student to report another for cheating…"
Such an exam system began in June 1921, according to a May 24, 1921 Daily article "Honor System Proposal Receives Approval and Support of Faculty". The first mention of something akin to the present Honor Code appears to be a January 23, 1929 Daily publication "The Honor System at Stanford", which explicitly calls out the responsibility of students to report on classmates who cheat on exams. Concerns about relying upon students reporting on each other appeared in the next academic year (e.g., see the May 6-7, 1930 articles "Harlow Rothert Calls Upon Student Body to Strengthen Honor Code" and "Students Sign Letter Pledging Enforcement of Honor Code: Fear Downfall of System") and have arisen regularly ever since.
The previous version of the Honor Code. The previous Honor Code said that: "students, individually and collectively...will do their share and take an active part in seeing to it that others as well as themselves uphold the spirit and letter of the Honor Code."
Data reveals that the Honor Code is not working as intended. It is widely recognized that the student reporting aspect of the Honor Code does not function as originally intended. As recent examples: during 2018-19 there were 136 Honor Code violations reported of which 2 came from students, during 2019-20 there were 191 Honor Code violations reported of which 0 came from students, and during 2020-21 (remote instruction) there were 393 Honor Code violations reported of which 0 came from students.
An early failure to ensure high levels of student engagement with the Honor Code was the "permanent Honor Committee" consisting of 15 students with annual elections, established on February 6, 1929 according to the Daily article "Permanent Committee for Honor Code Set Up by Executive Body". Unlike the long-running student Honor Committees at Princeton and University of Virginia, Stanford's "permanent" Honor Committee quickly disappeared: after an election on May 27, 1929, it is never again mentioned in the Daily.
To protect the university's reputation, the value of its degrees, and the integrity of the assessment system, sustainable action is now required. Concerns about academic integrity will likely only get worse if the university does not revisit its current policies and plan accordingly.
Creation of C-12 to address the problem. The above breakdown and the ever-changing interaction of technology with academic work were primary factors in the convening of the Committee of Ten (later Twelve) in 2019-20. The dramatic emergence of ChatGPT and its sequel GPT-4 since November 2022 has expanded the scope of these issues considerably (e.g., humanities coursework is now impacted by technology in ways that were inconceivable before November 2022).
Sources that inform C-12's recommendations. The recommendations below, addressing numerous facets of academic integrity in the 21st century, are informed by three main sources:
- feedback about academic integrity policies and procedures at 23 institutions (public and private) via a mix of written and oral communication;
- in-person and virtual discussions (pre and post ChatGPT) with students representing a cross section of the undergraduate and graduate student bodies;
- many virtual discussions (pre ChatGPT) with deans and faculty from all Schools at Stanford.
Proposals for further AIWG Investigation. In professional licensing (for doctors, lawyers, accountants, etc.) and nearly all high schools and post-secondary educational institutions (including many with Honor Codes), proctoring of exams is a standard and normal activity. Thus, Stanford students have experienced it to date (with great academic success) and most will experience it again at some point in their future education and careers. Undergraduate student outreach on in-person proctoring yielded slightly under half opposed to in-person proctoring during exams here, and the rest equally split between "yes" and "maybe".
In this appendix, further recommendations are given for matters that the AIWG should look into with the aim of formulating sustainable policy proposals. It may reassign any of these to a standing committee at Stanford with student and faculty membership that is well-suited and agreeable to taking it up. If the multi-year study of equitable proctoring practices determines that implicit bias is an issue (there is no evidence that students supervising each other, as expected under the current Honor Code, is any less prone to implicit bias) then implicit bias training can be incorporated into the state law-mandated sexual and gender-based harassment and abusive conduct training of faculty, lecturers, and teaching assistants.
- Forbid all timed take-home exams that are for a shorter time period within a longer one. This is the essentially unanimous opinion from student outreach, including the professional schools.
Further information. The word "exam" means work designed to be done continuously without interruption (apart from bathroom breaks, and exceptions such as for some OAE students) and without any communication with other humans (apart from seeking clarifications from course staff), and "take-home" means that the exam is to be done away from all supervision by course staff or designated alternatives (team coach, etc.). This does not apply to open-Internet exams to be worked on for however long a student wishes within a time period, such as 24 or 48 hours; academic integrity concerns for such exams are beyond the scope of these recommendations.
Rationale. Even if an exam is open-Internet, requiring students to work away from supervision by course staff or designated alternatives for a specified time duration within a longer one is a breeding ground for cheating or working on the exam for much longer than the official time duration (e.g., a student looks at a friend's copy of the exam before officially downloading their own copy, thereby circumventing time-stamping). All students recognize this reality, and it led to a huge increase in cheating during the remote instruction period.
Additional discussion. Currently SCPD students and traveling athletes on Stanford teams take exams (under the supervision of an exam monitor) for a fixed time period within a larger window, so AIWG's policy proposal on this matter will need to account for exceptions.
There are two (related) practical difficulties about which the university has been warned for many years:- the inadequate level of classroom space (as buildings with general-use classrooms are regularly knocked down and no comparable amount of general-use classroom space is built to replace what has been lost),
- the absence of sufficient resources to require -- as is done routinely at very many institutions from community colleges up through peer universities -- that OAE be 100% responsible for implementing everything it approves (compatibly with course requirements and policies; i.e., OAE cannot overrule those). The ever-increasing OAE related logistics must be 100% the responsibility of an OAE given adequate resources to do the job.
This highlights the urgency of addressing the institutional problems (1) and (2).
- Require course syllabi to be clear about collaboration and consultation.
Primary requirement. All course syllabi must provide clear guidance in writing on expectations for: citations, the use of permitted aid, and what is allowed in terms of modes of collaboration and resources for consultation in the coursework (homework, essays, projects, etc.). It is not realistic for the syllabus to provide an exhaustive list of all forms of unpermitted aid, but clear minimal parameters must be included. This guidance must be given at the start of the course.
Rationale. Due to shopping courses early in each quarter and for sustainability over time, the written syllabus constitutes the official record of Honor Code expectations in the course to which both students and course staff should refer (much as the syllabus provides the official record for other course policies; e.g., homework deadlines and the exam schedule).
A student may be enrolled in courses across a wide array of fields, and cannot be expected to know beforehand the standard conventions for academic discourse within each such field (nor is it reasonable to expect students to track down a list of departmental conventions that is not documented within the course material itself). Documenting course expectations in the written syllabus in a readable format is essential to prevent a minefield of “gotcha" concerns. Students have expressed frustration with having to keep track of varying expectations while enrolled in courses across many fields, particularly in the absence of sufficient clarity in the course material when expectations vary across different types of assignments.
Student duty to read the syllabus. Students are responsible for reading the syllabus, inquiring as to the use of aid beyond what is listed, and using common sense (e.g., using ChatGPT on an exam when it isn't expressly allowed, or copying a classmate's exam work, are forbidden).
The primary responsibility for educating students about the Honor Code resides with the university, but course staff are strongly encouraged to reinforce the importance of academic integrity.
Additional requirements and discussion- Such written guidance must be reasonably efficient (not go on for many pages and thereby be unrealistic to expect students to read) and should aim to achieve two goals:
- clear articulation of what the Honor Code means concretely in the course, via how it applies to coursework;
- education about why academic integrity is a value of the field of the course and/or what it means in the context of the course.
- When there are different expectations on different types of assignments (e.g., full collaboration is fine for one assignment, but on another you can work with a friend but not collaborate on the written submission, etc.), each assignment must come with a written reminder about these matters specific to the assignment.
- Syllabi should encourage asking questions about the use of aid when they arise. Some "common misconceptions" with aid are listed near the bottom of this OCS webpage on the Honor Code.
- To the extent course staff introduce assignments during class time, it is strongly encouraged that expectations on collaboration, citation, and so on be highlighted, especially when these differ across assignments.
- It is strongly recommended that instructors include a brief discussion on the first day of class about the role of the Honor Code in the course. That will help to improve the culture around academic integrity at Stanford by reinforcing our individual and collective commitment to the community standards as expressed in the Honor Code.
- Such written guidance must be reasonably efficient (not go on for many pages and thereby be unrealistic to expect students to read) and should aim to achieve two goals:
- Require academic integrity violations be reported.
Primary requirement. Course staff encountering an academic integrity violation should report it to OCS. All such OCS submissions beyond standard norms must have a basis in course policies clearly articulated in the written syllabus. Standard norms are field independent and universally recognized as cheating; e.g., looking at a classmate's exam work during an exam or consulting Internet resources during an exam that doesn't expressly allow them.
Rationale. The outreach to other institutions shows that there will always be some degree of non-reporting of integrity violation concerns, so the best that can be done is to reduce its frequency. The purpose of the requirement on submissions beyond standard norms is to be a sustainable mechanism to ensure that written syllabi provide clear guidance about the Honor Code in the context of the course (as is currently done in some departments but should be universal).
Additional requirements and discussion. If an Honor Code violation beyond standard norms is submitted without a basis in clear written guidance from course material (particularly for collaborative work) then it should be dismissed.
The revision of the Judicial Charter to have multiple levels (with an educational component for the lowest tier) and a more streamlined process should reduce the occurrence of instructors handling cases outside official channels. - Require Honor Code and Fundamental Standard education in ResEd training.
Primary requirement. Annual training (of 1 - 2 hours) on the Honor Code and Fundamental Standard for Residential Assistants, Community Associates, and other appropriate ResEd staff.
Rationale. This training will ensure that Residential Assistants and Community Associates can be a resource about the Honor Code and Fundamental Standard in student residences.
Additional requirements and discussion. The training must be provided to TA's as a mandatory part of their TA training (from the university or their own department or program) either before their work begins or within the first two weeks of their first TA quarter. It should be part of their weekly hours, and include a discussion of using such knowledge to guide students. - Require signing statement and quarterly confirmation of familiarity with the Honor Code.
Further information. Require that for every exam and major paper/project in a course, enrolled students sign a short statement attesting to their awareness of and commitment to abide by the Honor Code. Such a practice is already recommended guidance.
As is done at some other institutions, also require that when students (including Summer students) enroll in classes each quarter, they acknowledge and affirm their familiarity with the Honor Code and commitment to abide by it.
Rationale. The purpose is to improve the culture around the Honor Code, and ensure that lower-tier sanctions for a first violation do not lead to the impression that cheating until caught is an acceptable course of action.
Discussion. The signing statement may take on a variety of forms; e.g., the front page of an exam could contain the text of the Honor Code and the phrase "I reaffirm my commitment to adhere to Stanford's Honor Code throughout my time at Stanford, including during this exam," below which students are required to sign their name. Such signing statements are already common in many large introductory courses here, so this should not be too burdensome to implement (it is common practice at some other institutions with an Honor Code) nor entail increasing the amount of bureaucracy here. The OCS website should maintain examples of such statements (any of which instructors are welcome to use in whatever manner seems suitable) and include guidance for instructors and an FAQ for students and instructors.
The quarterly webform for the Honor Code affirmation can have a link to a short synopsis. - Require more training for international students on plagiarism norms in the United States.
Further information. For international students (at all levels), pre-matriculation online training related to plagiarism norms in this country has to be increased.
Rationale. Feedback from other institutions confirms plagiarism arises especially often with international students. Currently there is some such training that occurs at Stanford, but feedback within Stanford indicates that plagiarism cases for international students remain a significant concern and so the training needs to be expanded. - Improve scheduling of hearing panels.
Primary requirement. For panel hearings of Honor Code violations, schedule hearing windows in advance at some regular times.
Rationale. The use of standing panels has been effective at managing schedule challenges at other institutions, so it should improve scheduling logistics for panel hearings here. The current panel process here takes far too long compared with other institutions.
Additional requirements and discussion. Providing a modest stipend to student participants may help in recruitment efforts for such panels.
The investigative stage for Honor Code violations is done with efficiency at Harvard and Princeton using a small number of investigators, so Stanford should find out more about how their processes work (especially at Princeton, whose system is more similar to Stanford's in terms of the burden of proof) to allocate additional resources for investigators here. The investigation stage here often takes a long time, so emulating efficiencies at other institutions should help to reduce this time burden.
General observations. Here are additional ideas that emerged from feedback received in C-12 outreach.
- Regarding pro-active attention to impact of artificial intelligence on academic integrity
The university should designate a specific office (or offices coordinating across Schools) required to keep up to date on AI, produce annual reports on the matter, and develop both policy recommendations and sensible instructor training opportunities. - Regarding pro-active attention to discouraging academic dishonesty
Stanford must be pro-active about obtaining institutional licenses for software determined to be effective for academic dishonesty detection (e.g., plagiarism).
The Board on Conduct Affairs (BCA) recently approved requiring students to put their backpacks at a distance during exams (e.g., at the side of the room). This approach to reducing cheating via electronic devices should be explored for sustainable practices and be communicated to all academic departments and interdisciplinary programs. - Regarding the relationship of health to academic integrity
Develop sustainable policies to provide health insurance and mental health support to students (including graduate students) given a suspension sanction. The role of suspension in sanctions should be investigated to ensure its use is equitable in the consequences. - Regarding the obligation of Stanford University to underprepared admitted students
Providing substantially more resources to genuinely improve learning of course content among those admitted with underpreparedness must be an institutional priority. There has been progress in recent years, such as SOAR (for writing and math) and Math 18, but more needs to be done.
This is necessary to fulfill Stanford's promise (see the "Future of the Major" report, such as the start of Section 1 and Section 4.1 therein) that all admitted undergraduate students can successfully access all majors with 4-year degrees. The experienced instructors in many large introductory courses (especially in quantitative fields) can attest to increasing effects of underpreparedness: the resulting pressures can impact student choices related to both majors and academic integrity (though lack of academic preparedness is not an acceptable excuse for cheating). - Regarding communication about academic integrity processes
Webpages on the OCS website that spell out the academic integrity process should include a synopsis via a flowchart. This has been implemented at Baruch College, Caltech, UC Merced, and University of Virginia. - Regarding annual publication of Honor Code case summaries
For broad awareness about the current level of academic integrity violations, the OCS website should emulate the policy of some peer institutions to annually post anonymized summaries about Honor Code violations for public viewing in a visible place. This is done at Yale (look at the bottom of that webpage), Princeton, and University of Chicago. Currently the OCS annual reports on this are behind a firewall.
Concluding thoughts.
- For the coming years, the "3 books" tradition should incorporate a book touching upon themes of academic integrity. This can support a valuable discussion during New Student Orientation around academic integrity and related ethics, and restore a sense of student investment in the Honor Code. The most impactful orientation about the Honor Code or academic integrity occurs in small peer discussion, generally in student residences (e.g., guided by RA's, who should be trained to bring to life the realities of what Honor Code violations can entail).
Alternatively, Stanford can insert an academic integrity unit into a course for all undergraduates (e.g., in COLLEGE, or a mixture of that and Residential Advising discussion), to make the Honor Code a genuine and respected part of the student culture, not merely something to click through. One challenge with putting all significant discussion into a required course is that if a new undergraduate takes that course after the Fall then they are missing Honor Code training for Fall classes. - The BCA should convene an annual meeting of stakeholder groups from across the university (including the professional schools) to solicit feedback on the Honor Code in practice. Here, "stakeholder group" refers to a collection of departments and/or programs designated by Deans of Schools (e.g., the School of Humanities and Sciences has three clusters -- Humanities and Arts, Natural Sciences, and Social Sciences -- that could each be regarded as a stakeholder group) and to student government entities (e.g., ASSU, UGS, GSC).
Each stakeholder group should be invited to name representatives to attend. Departments or schools with a high incidence of violations in the previous year(s) should also be strongly encouraged to send a representative to address concerns specific to their students and faculty/lecturers. (Violation rate information can be provided to the BCA by OCS.) Recommendations from this meeting should be shared with OCS and the Faculty Senate, and may contribute to bylaw changes proposed by the BCA.
Representatives who are not students or members of OCS or OGC should all be members of instructional staff (e.g., faculty, lecturers) and/or student services or Academic Advising staff. This ensures feedback comes from those most directly involved in the university's academic mission. - At least once every five years, an Honor Code Review Committee should be convened with representation across all stakeholder groups (consisting of students and instructional staff along with at least one member from OCS and from OGC), and at least one representative from the BCA who has served for at least one academic year on the board. This Committee will undertake an in-depth review of the Honor Code to assess its form and function in the Stanford community.
Work done by that Committee may include: convening focus groups and town halls, reviewing relevant data, and checking if the Honor Code remains relevant and functional (in light of developing technology, etc.). In particular, it should review collaborative work in courses from fields with an excessive number of Honor Code violations. Such a review should check that a) syllabi incorporate reference to the Honor Code, b) the written policies regarding permitted aid for the course are expressly included and clear, and c) when permitted aid varies by assignment, the syllabus clearly articulates the aid permitted for each assignment.
This review should produce a report provided to the OCS and/or BCA, including recommendations as appropriate. In particular, the review should identify resources related to academic integrity that need improvement or availability, and determine the necessary and sustainable level of investment of university resources. The report must be posted on a public-facing OCS website and its main findings should be shared (e.g., by the BCA) with the Faculty Senate and ASSU.
