Ethical Reasons to Avoid Using AI Apps for Student Assessment

September 10, 2024

It’s the start of a new school year here in North America. We are into the second week of classes and already I am hearing from administrators in both K-12 and higher education institutions who are frustrated with educators who have turned to ChatGPT and other publicly-available Gen AI apps to help them assess student learning.

Although customized AI apps designed specifically to assist with the assessment of student learning already exist, many educators do not yet have access to such tools. Instead, I am hearing about educators turning to large language models (LLMs) like ChatGPT to help them provide formative or summative assessment of students’ work. There are some good reasons not to avoid using ChatGPT or other LLMs to assess student learning.

I expect that not everyone will agree with these points, please take them with the spirit in which they are intended, which to provide guidance on ethical ways to teach, learn, and assess students’ work.

8 Tips on Why Educators Should Avoid Using AI Apps to Help with Assessment of Student Learning

Intellectual Property

In Canada at least, a student’s work is their intellectual property. Unless you have permission to use it outside of class, then avoid doing so. The bottom line here is that student’s intellectual work is not yours to share to a large-language model (LLM) or any other third party application, with out their knowledge and consent.

Privacy

A student’s personal data, including their name, ID number and other details should never be uploaded to an external app without consent. One reason for this blog post is to respond to stories I am hearing about educators uploading entire student essays or assignments, including the cover page with all the identifying information, to a third-party GenAI app.

Data security

Content uploaded to an AI tool may be added to its database and used to train the tool. Uploading student assignments to GenAI apps for feedback poses several data security risks. These include potential breaches of data storage systems, privacy violations through sharing sensitive student information, and intellectual property concerns. Inadequate access controls or encryption could allow unauthorized access to student work. 

AI model vulnerabilities might enable data extraction, while unintended leakage could occur through the AI app’s responses. If the educator’s account is compromised, it could expose all of the uploaded assignments. The app’s policies may permit third-party data sharing, and long-term data persistence in backups or training sets could extend the risk timeline. Also, there may be legal and regulatory issues around sharing student data, especially for minors, without proper consent.

Bias

AI apps are known to be biased. Feedback generated by an AI app can be biased, unfair, and even racist. To learn more check out this article published in Nature. AI models can perpetuate existing biases present in their training data, which may not represent diverse student populations adequately. Apps might favour certain writing styles (e.g., standard American English), cultural references, or modes of expression, disadvantaging students from different backgrounds. 

Furthermore, the AI’s feedback could be inconsistent across similar submissions or fail to account for individual student progress and needs. Additionally, the app may not fully grasp nuanced or creative approaches, leading to standardized feedback that discourages unique thinking.

Lack of context

An AI app does not know your student like you do. Although GenAI tools can offer quick assessments and feedback, they often lack the nuanced understanding of a student’s unique context, learning style, and emotional or physical well-being. Overreliance on AI-generated feedback might lead to generic responses, diminishing the personal connection and meaningful interaction that educators provide, which are vital for effective learning.

Impersonal

AI apps can provide generic feedback, but as an educator, you can personalize feedback to help the student grow. AI apps can provide generic feedback but may not help to scaffold a student’s learning. Personalized feedback is crucial, as it fosters individual student growth, enhances understanding, and encourages engagement with the material. Tailoring feedback to specific strengths and weaknesses helps students recognize their progress and areas needing improvement. In turn, this helps to build their confidence and motivation. 

Academic Integrity

Educators model ethical behaviour, this includes transparent and fair assessment. If you are using tech tools to assess student learning, it is important to be transparent about it. In this post, I write more about how and why deceptive and covert assessment tactics are unacceptable.

Your Employee Responsibilities

If your job description includes assessing student work , you may be violating your employment contract if you offload assessment to an AI app.

Concluding Thoughts

Unless your employer has explicitly given you permission to use AI apps for assessing student work then, at least for now, consider providing feedback and assessment in the ways expected by your employer. If we do not want students to use AI apps to take shortcuts, then it is up to us as educators to model the behavior we expect from students.

I understand that educators have excessive and exhausting workloads. I appreciate that we have more items on our To Do Lists than is reasonable. I totally get it that we may look for shortcuts and ways to reduce our workload. The reality is that although Gen AI may have the capability to help with certain tasks, not all employers have endorsed their use in same way.

Not all institutions or schools have artificial intelligence policies or guideline, so when in doubt, ask your supervisor if you are not sure about the expectations. Again, there is a parallel here with student conduct. If we expect students to avoid using AI apps unless we make it explicit that it is OK, then the same goes for educators. Avoid using unauthorized tech tools for assessment without the boss knowing about it.

I am not suggesting that Gen AI apps don’t have the capability to assist with AI, but I am suggesting that many educational institutions have not yet approved the use of such apps for use in the workplace. Trust me, when there are Gen AI apps to help with the heaviest aspects of our workload as educators, I’ll be at the front of the line to use them. In the meantime, there’s a balance to be struck between what AI can do and what one’s employer may permit us to use AI for. It’s important to know the difference — and to protect your livelihood.

Related post:

The Use of AI-Detection Tools in the Assessment of Student Work https://drsaraheaton.wordpress.com/2023/05/06/the-use-of-ai-detection-tools-in-the-assessment-of-student-work/

____________________________

Share this post:
Ethical Reasons to Avoid Using AI Apps for Student Assessment – https://drsaraheaton.com/2024/09/10/ethical-reasons-to-avoid-using-ai-apps-for-student-assessment/

This blog has had over 3.6 million views thanks to readers like you. If you enjoyed this post, please “like” it or share it on social media. Thanks!

Sarah Elaine Eaton, PhD, is a faculty member in the Werklund School of Education at the University of Calgary, Canada. Opinions are my own and do not represent those of my employer.

Sarah Elaine Eaton, PhD, Editor-in-Chief, International Journal for Educational Integrity


Academic integrity and artificial intelligence in higher education (HE) contexts: A rapid scoping review

September 4, 2024

In this post, I’d like to give a shoutout to Beatriz Moya, who led a rapid review on academic integrity and artificial intelligence.

A screenshot of a title page of an academic article. There is purple and black text on a white background.
Title page of “Academic Integrity and artificial intelligence in higher education (HE) contexts: A rapid scoping review”.

Here is the reference:

Moya, B. A., Eaton, S. E., Pethrick, H., Hayden, A. K., Brennan, R., Wiens, J., & McDermott, B. (2024). Academic integrity and artificial intelligence in higher education (HE) contexts: A rapid scoping review. Canadian Perspectives on Academic Integrity, 7(3). https://doi.org/10.55016/ojs/cpai.v7i3

Abstract

Artificial intelligence (AI) developments challenge higher education institutions’ teaching, learning, assessment, and research practices. To contribute evidence-based recommendations for upholding academic integrity, we conducted a rapid scoping review focusing on what is known about academic integrity and AI in higher education before the emergence of ChatGPT. We followed the Updated Reviewer Manual for Scoping Reviews from the Joanna Briggs Institute (JBI) and the Preferred Reporting Items for Systematic reviews Meta-Analysis for Scoping Reviews (PRISMA-ScR) reporting standards. Five databases were searched, and the eligibility criteria included higher education stakeholders of any age and gender engaged with AI in the context of academic integrity from 2007 through November 2022 and available in English. The search retrieved 2,223 records, of which 14 publications with mixed methods, qualitative, quantitative, randomized controlled trials, and text and opinion studies met the inclusion criteria. The results showed bounded and unbounded ethical implications of AI. Perspectives included: AI for cheating; AI as legitimate support; an equity, diversity, and inclusion lens into AI; and emerging recommendations to tackle AI implications in higher education. The evidence from the sources provides guidance that can inform educational stakeholders in decision-making processes for AI integration, in the analysis of misconduct cases involving AI, and in the exploration of AI as legitimate assistance. Likewise, this rapid scoping review signals possibilities for future research, which we explore in our discussion.

Keywords

academic integrity, artificial intelligence, academic misconduct, higher education, rapid scoping review, large language models (LLM)

This is a fully open access article. You can download a copy of the full article here: https://doi.org/10.55016/ojs/cpai.v7i3

Related posts:

Exploring the Contemporary Intersections of Artificial Intelligence and Academic Integrity https://drsaraheaton.wordpress.com/2022/05/17/exploring-the-contemporary-intersections-of-artificial-intelligence-and-academic-integrity/

New project: Artificial Intelligence and Academic Integrity: The Ethics of Teaching and Learning with Algorithmic Writing Technologieshttps://drsaraheaton.wordpress.com/2022/04/19/new-project-artificial-intelligence-and-academic-integrity-the-ethics-of-teaching-and-learning-with-algorithmic-writing-technologies/

The Use of AI-Detection Tools in the Assessment of Student Workhttps://drsaraheaton.wordpress.com/2023/05/06/the-use-of-ai-detection-tools-in-the-assessment-of-student-work/

____________________________

Share this post: Academic integrity and artificial intelligence in higher education (HE) contexts: A rapid scoping review – https://drsaraheaton.com/2024/09/04/academic-integrity-and-artificial-intelligence-in-higher-education-he-contexts-a-rapid-scoping-review/

This blog has had over 3.6 million views thanks to readers like you. If you enjoyed this post, please “like” it or share it on social media. Thanks!

Sarah Elaine Eaton, PhD, is a faculty member in the Werklund School of Education at the University of Calgary, Canada. Opinions are my own and do not represent those of my employer.

Sarah Elaine Eaton, PhD, Editor-in-Chief, International Journal for Educational Integrity


Invitation to Participate: Research Study on Artificial Intelligence and Academic Integrity: 

April 19, 2023

The Ethics of Teaching and Learning with Algorithmic Writing Technologies 

On the right there is a black robotic hand and forearm. On the left there is a human hand and forearm. The forearm is tatooed. One finger from each hand is touching the other.
Photo by cottonbro studio on Pexels.com

Academic misconduct has taken various forms in present-day educational systems. One method that is on the rise is the use of artificially generated software compositions. The capabilities and sophistication of these new technologies are improving steadily. We are conducting a study to gauge the sophistication of the current artificial intelligence (AI) software-generated text. To that end, we are recruiting participants to evaluate the level of writing level of small compositions (260 words in length at most).

Your participation in this study would be to evaluate two small pieces of text presented in a survey and optionally make comments on your observation. We appreciate your consideration in this matter. This research provides an opportunity for the participants to contribute to the state of AI software used for various educational purposes. Participation in this study is voluntary, and you are free to terminate the survey and withdraw at any time and for any reason without censor. There are no known physical, psychological, or social risks associated with participation in the study.

All demographic data collected will be kept strictly confidential. Only the researchers listed in this letter will have access to the raw data. The data (in electronic format) will be retained indefinitely. Participation in the study will be asked for some basic demographic information and then presented with a 260- word length composition. After reading, the participants will be asked to evaluate the level, assign a mark to the composition, and note any pertinent observations. The second piece of composition, also of the same length, will be followed by the same set of questions. The total anticipated time for completing the survey is about 9-12 minutes, but it can vary based on reading speed and consideration afforded to the assigned grade.

If you have any questions or concerns about your participation in this study, you can contact the Principal Investigator, Dr. Sarah Elaine Eaton, seaton (at) ucalgary.ca

This study is funded by a University of Calgary Teaching and Learning Grant. This study has been approved by the Conjoint Faculties Research Ethics Board at the University of Calgary: REB22-0137.

To take the survey, click here.

_________________________________

Share this post: Invitation to Participate: Research Study on Artificial Intelligence and Academic Integrity: The Ethics of Teaching and Learning with Algorithmic Writing Technologies – https://wp.me/pNAh3-2U3

This blog has had over 3 million views thanks to readers like you. If you enjoyed this post, please “like” it or share it on social media. Thanks! Sarah Elaine Eaton, PhD, is a faculty member in the Werklund School of Education, and the Educational Leader in Residence, Academic Integrity, University of Calgary, Canada. Opinions are my own and do not represent those of the University of Calgary.

 


Book launch: Fake Degrees and Fraudulent Credentials in Higher Education

March 7, 2023

Carleton University Innovation Hub is pleased to host this public event.

Fake Degrees and Fraudulent Credentials in Higher Education

Join editors/authors Sarah Elaine Eaton, Jamie J. Carmichael, and Helen Pethrick in the Innovation Hub on Friday March 24, 2023 for the launch of their new book Fake Degrees and Fraudulent Credentials in Higher Education. This hybrid event will feature reading from the book and continue an important conversation facing many people in the world today.

Event Date: March 24th, 11:00 am – 12:00 (Eastern Standard Time)  

Event Location: Innovation Hub, 2020 Nicol Building, Carleton University or online

Hybrid Option:  A zoom Link will be provided via email to event registrants.

For more information or to register – https://carleton.ca/innovationhub/book-launch/

Related posts

_________________________________

Share or Tweet this: Book launch: Fake Degrees and Fraudulent Credentials in Higher Education https://wp.me/pNAh3-2SD

This blog has had over 3 million views thanks to readers like you. If you enjoyed this post, please “like” it or share it on social media. Thanks!

Sarah Elaine Eaton, PhD, is a faculty member in the Werklund School of Education, and the Educational Leader in Residence, Academic Integrity, University of Calgary, Canada. Opinions are my own and do not represent those of the University of Calgary.


Fake Degrees and Fraudulent Credentials in Higher Education: A Synopsis of the Book

January 28, 2023

Fake Degrees and Fraudulent Credentials in Higher Education was published by Springer Nature in January, 2023. It features over a dozen chapters on various topics related the broad theme of credential fraud.

The introductory chapter includes an overview and historical perspectives of key issues. In this opening chapter, the authors connect the dots between the industries that supply fake degrees and fraudulent academic documents (including but not limited to bogus reference letters and tampered transcripts) to the contract cheating industries (e.g., term paper mills, student proxy services, and examination impersonators who take tests on behalf students) and to the admissions fraud industry. In addition, a connection is made to the scholarly paper mill industry which includes various violations of scientific publication ethics including fake data, fabricated scholarly articles and authorship for sale.

We present these connected industries through our Ecosystem of Commercial Academic Fraud model (Eaton & Carmichael, 2022). They synthesize what is known about the size and scope of the industry, estimating its valuation to be at least $21 Billion USD.

In Chapter Two, Jamie Carmichael and I (who are Canadian) argue that Canada is vulnerable to admission and credential fraud. This argument will be substantiated through (i) a survey that spanned coast-to-coast within Canada, (ii) media review that targeted the response to Operation Varsity Blues in Canadian newspapers, and (iii) a comparative analysis with another survey on this topic to gauge how Canada is fairing compared to others across the globe. This data triangulation resulted in an in-depth examination of admissions fraud in Canada, with 14 recommendations for practitioners, researchers, and those involved in policy reform.

In the third chapter, FBI Special Agent (retired), Allen Ezell, takes readers on a tour of “Axact, the world’s largest diploma mill”. Ezell has more than 40 years of experience investigating “fake high schools, colleges, universities, and counterfeiters” (p. 53). He writes that, “Axact is a classic example of a criminal enterprise. Even its own employees refer to it as the tower of frauds and house of lies.” (p. 49). This chapter is the longest in the book by far, spanning 49 pages, complete with concrete details, photos, and insider information not available anywhere else. Ezell explains that diploma mills, “are professional operations that take planning, preparation, and organization to run smoothly. Like legitimate businesses, they have business models, conduct market studies, have financial forecasts, perform cost analyses, set daily/ weekly/monthly sales goals, and offer sales incentives. They constantly survey their competition, and actions by law enforcement and regulators, to determine the direction the wind is blowing.” (p. 53)

Chapter Four, by Joanne Duklas, addresses how “digitization and technology have improved electronic exchange practices in the areas of document and data management and reduced occurrences of fraud thereby encouraging greater trust in the verification and assessment process for admission and transfer” (p. 95). Duklas discusses how background checks regularly show evidence of misrepresentation and fraud. Duklas’s in-depth exploration of  key issues related to electronic exchange processes. She provides details about Canada’s national document exchange network, MyCreds™ | MesCertif™, launched in 2020. She notes that, “Building solid bridges (technical infrastructure, standards, legislation, policies, and procedures) between issuers and receivers of official documents remains important for a trusted, quality assured ecosystem. Various technology solutions are solving credential fraud and creating greater trust in the chain of custody and bone fides of official documents.” (p. 110)

In Chapter 5, Kirsten Hextrum “considers how legal athletic admissions mirror the largest college admission conspiracy in US history: Operation Varsity Blues (OVB)” (p.  115). Hextrum analyzed 1487 college athletes’ demographic data; 47 life- history interviews with college athletes; and admission-related documents. Her results reveal how athletes invest to develop athletic talent, disputing college leaders’ claims that athletic admissions create diverse cohorts. She addresses important issues related to equity, diversity, and inclusion, as her findings show that “athletic investments create homogenous cohorts as White, middle-class youth are overrepresented as college athletes.” (p. 115). In addition, Hextrum discusses, “how colleges authenticate athletes’ credentials” (p. 115), finding that “universities use inconsistent and arbitrary measures—sometimes admitting athletes with little sport experience” (p. 115). Hextrum’s findings indicate that “athletic admissions misalign with the public’s interest because they are fixed to favour White, middle-class athletes and remain vulnerable to fraud” (p. 115).

In the next chapter, Stella-Maris Orim and Irene Glendinning, write about “Corruption in Admissions, Recruitment, Qualifications and Credentials” from the perspective of quality assurance. They draw on research conducted in 2017–2018 for the Council of Higher Education Accreditation’s International Quality Group (CIQG) “into how Accreditation and Quality Assurance Bodies (AQABs) respond to corruption” (p. 133). They address the question: Who is responsible for reducing corruption in education and research? (p. 133). They present findings from their empirical research, leading them to their evidence-informed conclusion that “higher education providers around the world are aware of corruption, fraud and malpractice in student assessment.”, but that “that much less attention has been given by institutions to the types of integrity breaches that we have looked at in this chapter, fraud in admissions and recruitment and in credentials and qualifications” (p. 145).

Chapter 7, by Özgür Çelik and Salim Razı, addresses favouritism and professorial recruitment practices in Turkish higher education institutions. They address the issue of fraud and corruption indirectly, through an analysis of 66 news stories that address favourtism in Turkish universities. They discuss how favouritism erodes public trust, noting that “the negative consequences of favouritism are far-reaching” (p. 154). Their findings showed that nepotism, cronyism, and patronage, were key areas of concern. They found that when specific individuals were identified for academic positions that customized job descriptions can be written so that only that particular individual could be deemed qualified. Although their chapter focuses specifically on Turkey, there are lessons that are transferrable to other countries with regards to corrupt hiring practices in higher education and other sectors.

The next two chapters address fraud in standardized English-language proficiency testing. Soroush Sabbaghan and Ismaeil Fazel’s chapter aims to “shed light on the complexities and the apparent disconnect between equity, integrity, fairness, and justice in standardized language proficiency tests and the integrity issues that can arise as a result” (p. 169). They point out that “at their core, standardized procedures imposed by testing centers ignore the fact that test-takers come from different socio-economic and sociocultural backgrounds with different interests, motivations and experiences of learning and using English” (p. 171). For those interested in exploring issues related to equity, diversity, and inclusion of international students, this chapter is a must read.

Angela Clark continues the discussion of fraud and corruption in English-language proficiency exams in her chapter that explores fraudulent test scores. Clark argues that “relying on a single language proficiency test score to determine an individual’s readiness is problematic, and also problematic is the lack of related academic research and data to help guide admissions decision-making” (p. 187). Clark presents concrete recommendations for “institutional stakeholders with ways to become better informed about these tests and their impact and approaches to help international NNES [non-native English-speaker] students succeed within their new academic disciplines and new academic culture” (p. 187) that include reconsidering admission criteria (p. 199) and instituting a post-entry language assessment (PELA) (p. 199).

In Chapter Ten, Brendan DeCoster uses systems theory to address admissions fraud. DeCoster explores the “culture of admissions fraud” in the United States. He proposes “a five-level framework for analyzing admissions fraud, noting how individuals, small groups and firms, larger firms, institutions, and political entities all play roles in establishing definitions of fraud, investigating fraud, prosecuting fraud, committing fraud, abetting fraud, and countering fraud” (p. 209). The five levels include micro, metaxy, meso, macro, and acro, and he provides examples for each level in his chapter.

Next, Jamie Carmichael leads a chapter on topic modelling, “an unsupervised machine learning technique commonly used in computer science as a research method” (p. 227). In this novel study, “data from 30 websites selling fake degrees were manually scraped, observations noted, and a topic model was built to identify risks within the dataset”, demonstrating that “that topic modeling can identify security risks by providing an environmental scan of the threat.” (p. 227). Twenty evidence-based recommendations are offered for higher education security professionals, senior leaders, and researchers.

In the penultimate chapter, together with Jamie Carmichael, I explore “what can happen when professor and educational leaders have fake or fraudulent degrees or other qualifications. We present four key issues: (a) the threat to institutional reputation; (b) the threat to the credentials awarded by the institutions; (c) the impact on students; and (d) material costs to the organization. Then, we propose seven recommendations to prevent or address academic qualification fraud: (a) verify applicant credentials; (b) develop or update internal risk assessment plans; (c) conduct an internal qualifications audit; (d) develop or update institutional codes of conduct; (e) develop an internal process to investigate allegations of credential fraud; (f) develop and follow internal quality assurance processes for courses, programs, and curricula; and (g) Develop or update crisis communications plans to include credential fakery or fraud. We conclude by emphasizing that moral outrage will not solve the problem of academic credential fraud.” (p. 251).

In our final chapter, we summarize key findings from the book, grouping them into “seven main categories: (a) historical perspectives and terminology; (b) a trend of global indifference; (c) criminal enterprises and security; (d) fraud in standardized language proficiency testing; (e) athletic credentialism; (f) the role of the institution; (g) hiring, and (h) technology. We discuss the significance and limitations of the book, concluding with calls to action for more research and resources to better understand and address the growing problem of the ecosystem of academic fraud that continues to grow” (p. 269).

This book was a passion project that we undertook during COVID-19. We hope it is useful to others who are dedicated to upholding academic integrity and raising awareness about the threats posed to higher education by fake degrees, fraud, corruption, and quackery.

Related posts

Just published! Fake Degrees and Fraudulent Credentials in Higher Education

Fake Degrees and Fraudulent Credentials: Research Project Update

Scholarships Without Scruples: 3 Signs of Bogus Scholarships and Scams

Why Universities and Colleges Need Clear Policies to Deal with Fake COVID-19 Vaccination Records and Test Results

Degrees of Deceit: A Webinar

_________________________________

Share or Tweet this: Fake Degrees and Fraudulent Credentials in Higher Education: A Synopsis of the Book https://drsaraheaton.wordpress.com/2023/01/28/fake-degrees-and-fraudulent-credentials-in-higher-education-a-synopsis-of-the-book/

This blog has had over 3 million views thanks to readers like you. If you enjoyed this post, please “like” it or share it on social media. Thanks!

Sarah Elaine Eaton, PhD, is a faculty member in the Werklund School of Education, and the Educational Leader in Residence, Academic Integrity, University of Calgary, Canada. Opinions are my own and do not represent those of the University of Calgary.