College responsibilities in the context of public examinations of coursework

By Connor Berry, pedagogy leader - Greenhead Sixth Form College and Cressida Tweed, head of philosophy and RE - Woodhouse College

Introduction

A recent survey conducted in a sixth form of over 2,000 students highlights the increasing use of AI tools by students: compared to only a handful of teachers (9 out of 66), a significant majority of students (72%) reported using AI for some form of college work in the Spring Term.

AI use in education is rapidly growing. With major tech companies integrating AI into everyday platforms, student access and expertise will likely continue to increase.

As educators, we need to stay informed about the latest developments in AI to effectively identify potential misuse during exams and coursework. Both the DfE and Joint Council for Qualifications (JCQ) have recently updated their guidance to reflect the advances in AI and this section will explain how this can be applied in your setting.

a. Tackling the plagiarism problem

JCQ states: “While the potential for student artificial intelligence (AI) misuse is new, most of the ways to prevent its misuse and mitigate the associated risks are not; centres will already have established measures in place to ensure that students are aware of the importance of submitting their own independent work for assessment and for identifying potential malpractice.”

Plagiarism is not just an issue for NEA; it also affects some public examinations, in particular entrance examinations to universities. In some cases, students are able to take these at home (such as Cambridge's Philosophy and Foundation Year Admissions Assessments); while those tests are done under timed conditions, there is nothing preventing students from using or accessing AI (be it large language models or syntax checkers such as Grammarly).

The JCQ guidance on AI makes it clear that centres have a duty to train staff in relation to plagiarism and AI use, must update their plagiarism policy so that it includes the use of AI, and give clear guidance to students on what constitutes malpractice. Students complete their NEA for A levels, extended project qualifications, and public exams, all of which can be impacted by AI misuse.

The JCQ guidance on Al makes it clear that centres have a duty to train staff in relation to plagiarism and Al use, must update their plagiarism policy so that it includes the use of Al, and give clear guidance to students on what constitutes malpractice. Students complete their NEA for A levels, extended projects, and vocational qualifications, and sit entrance examinations with a varying degree of supervision depending on the institution, so you will need to be conscious of how your context makes Al use more or less possible.

The latest guidance from Ofqual, published in April 2024, focuses on the concerns AI raises in relation to:

• Ensuring fairness for students
• Maintaining validity of qualifications
• Protecting security
• Maintaining public confidence
• Enabling innovation

While centres have a responsibility to prevent plagiarism and train students and staff to use AI responsibly, Ofqual makes clear in their rules and guidance to awarding organisations that they must act on suspected malpractice. Unfortunately, the use of AI in NEA puts the integrity of qualifications at risk. This examination series (summer 2024), external moderators have been given specific training and guidance on recognizing instances of malpractice.

This includes looking out for:

• Sentence structure and length which is not consistent
• Use of vocabulary which falls outside of the usual terminology used in the qualification
• Varying quality of work/varying skillsets within one body of work
• Comparison with previous work
• Lack of references and direct quotes and citation
• Repetitive phrases or words
• Lack of clear judgment
• Lack of graphs and data tables where these would normally be expected

b. Checking for AI usage

No plagiarism software is fully reliable in detecting AI, and can discriminate against students for whom English is not the first language. Turnitin offers plagiarism software, but it has to be rolled out at institution-level and monitors all communication and data on college systems. It is also more effective on peer-to-peer plagiarism and traditional plagiarism than generative AI or use of writing assistance tools. Some applications are also able to change texts and rephrase paragraphs to hide use of AI; this can include adding syntax errors or misspelt words.

c. Signing off with confidence

Certain steps can be taken by institutions to mitigate the risks posed by AI in relation to plagiarism:

• Getting students to do some of their NEA under supervision in a college setting.
• Asking students to submit drafts at regular intervals. This also allows centres to provide evidence to examination boards should a suspected case of malpractice arise.
• Alternatively, asking students to make their drafts shareable so that teachers can monitor changes to drafts in the document history.
• Offering one-to-one support to students, including asking them to explain a paragraph or a key section of their coursework.
• Running an assembly explaining to the students the risks they are taking in using AI in NEA, explaining what malpractice entails and what will happen if they are found guilty of malpractice. The JCQ guidance offers examples of students who were disqualified because of malpractice.
• While students have to sign a candidate record form to confirm it is their own work, it might also be useful to ask students to sign a form specific to AI whereby they acknowledge that they have received training on this, understand the need for referencing their work (including use of AI), and understand that they may be disqualified if they have not done so.
• Telling students that the institution may/will use plagiarism detection software (and getting them to fill a consent form) can prove to be an effective deterrent even if you do not use a tool to detect AI.
• Alternatively, students could sign a student charter as part of the code of conduct when they enrol explaining their responsibilities as students.

Consent form for the use of plagiarism detection software to check my extended project

LAST NAME:
FIRST NAME:
STUDENT NUMBER:
COURSE:
TITLE:

JCQ regulations on coursework

• JCQ regulations state that: "the work which you submit for assessment must be your own"; "you must not copy from someone else or allow another candidate to copy from you."
  If you use the same wording as a published source, you must place quotation marks around the passage and state where it came from. This is called "referencing." You must make sure that you give detailed references for everything in your work which is not in your own words. A reference from a printed book or journal should show the name of the author, the year of publication, and the page number, for example: (Morrison, 2000, p29).
• For material taken from the internet, your reference should show the date when the material was downloaded and must show the precise web page, not the search engine used to locate it. This can be copied from the address line. For example: http://news.bbc.co.uk/onthisday/hi/dates/stories/october/28/newsid_2621000/2621915.stm, downloaded 5 February 2024.
• Where computer-generated content has been used (such as an AI chatbot), your reference must show the name of the AI bot used and should show the date the content was generated. For example: ChatGPT 3.5 (https://openai.com/blog/chatgpt/), 25/01/2024. You should retain a copy of the computer-generated content for reference and authentication purposes.
• You are required to produce a bibliography at the end of your work. This must list the full details of publications you have used in your research, even where these are not directly referred to, for example: Curran, J. Mass Media and Society (Hodder Arnold, 2005).

If you copy the words, ideas, or outputs of others and do not show your sources in references and a bibliography, this will be considered as cheating.

What is plagiarism?

Plagiarism is defined as submitting someone else's work or ideas as your own without a complete indication of the source. It includes copying word by word without acknowledgment of the source, but also borrowing text structures (e.g., line of argumentation or outline) and texts translated from a foreign language.

Use of plagiarism detection software

The extended project department may use plagiarism software to check each submitted extended project for plagiarism. For that purpose, your extended project will be electronically forwarded to a software service provider who will check for potential matches between the submitted work and work from other sources. For future comparisons with other projects, your project will be permanently stored in a database. Only Woodhouse College is allowed to access your stored project. You agree that your submission may be stored and reproduced only for the purpose of plagiarism assessment. Your examiner will be advised on the outcome of the plagiarism assessment.

Sanctions

• Each case of plagiarism constitutes an attempt to deceive in terms of the examination regulations. The exam board, Edexcel, will be informed if the department finds evidence of plagiarism.
• I confirm that I have read and understood the information in this document. I agree to the outlined procedure for plagiarism assessment and potential sanctions.

SIGNATURE OF THE STUDENT:
DATE:

d. Further discussion of DfE and JCQ guidance

In October 2023, the DfE released its guidance document titled "Generative artificial intelligence (AI) in education," already referenced in this article. It sets out the position of the Government in relation to generative AI but gives educators guidance on how to use AI effectively in its current form and to be aware of some of the limitations.

According to the DfE guidance, here's how AI can be a game-changer in education:

• Time-Saving Superhero: Generative AI can automate repetitive tasks like creating quizzes, worksheets, and basic lesson plans. This frees up valuable time for teachers to focus on what they do best: inspiring their students.
• Personalised Learning: Generative AI can tailor learning experiences to individual student needs. Imagine practice problems that adjust difficulty based on a student's progress, or summaries written in a student-friendly format.
• Engaging Content Creator: Imagine bringing history to life with an AI-generated simulation of a historical event or letting students explore scientific concepts through interactive AI-powered models.

With any new technology, there are things to consider:

• Fact-Checking Friend: Just like any research tool, it's important to critically evaluate the information AI generates. Teachers can guide students on how to check for accuracy and identify potential bias.
• Teach Them How: As more AI tools become available, it's important to teach students how to use them responsibly.

Key takeaways from the DfE guidance

• Focus on the Benefits: AI can be a powerful tool to enhance learning and empower teachers when used effectively.
• Be Responsible: Protect personal and special category data in accordance with data protection legislation as well as following copyright laws.
• Know the Limits: Use AI tools thoughtfully and critically evaluate the information they produce.
• Support: The Department for Education (DfE) is there to support teachers on this exciting journey.

The Joint Council for Qualifications (JCQ) has established guidelines on the proper use of AI in education. These were updated in February 2024.

Understanding these guidelines empowers teachers in two key ways:

• Confidence in Identifying Misuse: Clear guidelines equip us to effectively spot instances where AI is used inappropriately.
• Empowering Positive Use: Knowledge of the guidelines allows us to have productive discussions with students about leveraging AI responsibly within acceptable boundaries.

It's important to remember that your existing teaching skills and experience are invaluable in preventing AI misuse. Many schools and colleges already have measures in place to ensure academic integrity is maintained. The JCQ guidelines simply provide an additional framework to build upon your existing expertise.

By familiarising yourself with the JCQ guidelines, you can ensure a balanced approach: effectively preventing misuse while also exploring the potential positive applications of AI in the classroom.

The key features of the JCQ guidelines are:

• Remind students that any content produced using AI must be referenced and cannot be given marks — and a failure to reference the use of AI is malpractice.
• Make sure students and parents know that misusing AI is cheating and a form of malpractice.
• Stress the importance of the candidate declaration (which references AI use) when they submit their work for assessment.
• Where there has been no AI misuse: "When marking student work in which AI use has been acknowledged, and there are no concerns of AI misuse, the assessor must still ensure that if the student has used AI tools such that they have not independently met the marking criteria, they are not rewarded."
• On marking work: "Centres may determine, after careful consideration of any data privacy concerns, whether it is appropriate for their teachers and assessors to use AI tools to help mark student work. Where centres do permit AI tools to be used to mark student work, an AI tool cannot be the sole marker. A human assessor must review all of the work in its entirety and determine the mark they feel it warrants, regardless of the outcomes of an AI tool. The assessor remains responsible for the mark/grade awarded."

JCQ has produced some handy resources to assist teachers and students in understanding the rules and regulations when it comes to exams. Student-friendly posters can be printed on A3 and stuck up in classrooms where NEA/EPQ is delivered, and teachers should make reference to them and explain the guidance when delivering content to students.

There are also two very well-put-together and easy-to-navigate PowerPoint presentations designed by JCQ for delivering to staff and students. Use this link here to access them.

Useful resources

The National Extension College offers a free guide to AI to help learners understand what AI is and what constitutes malpractice.

A Guide to AI - National Extension College (nec.ac.uk)

AI Course for Educators - AI for Education