On this webpage, we discuss some of these challenges of assessing the knowledge of students who use English as an additional language (EAL) in the age of AI and steps instructors can take to mitigate them. We start with handwriting (in-class exams, bluebooks), and then move to oral assessment, class participation, and computer-assisted writing.
(We have created a mirror webpage for students, so that they can best prepare for each of these types of assessment.)
Handwriting (in-class exams, blue books, in-class essay writing)
Why use this form of assessment: To avoid the problem of AI-impacted writing, some faculty are using in-class writing as a form of assessment (e.g. blue books, in-class written exams, in-class essay writing) (Hunter, 2026). These forms of assessment retain some of the valuable aspects of computer-assisted writing. However, there are features of in-class writing that may negatively impact EAL students (Barkaoui & Knouzi, 2018; Lee, 2004).
The challenges with this form of assessment: The first challenge is one that would impact all students: writing by hand, without the assistance of a computer, is to write in an artificial environment. Today’s students were raised to write on laptops and other devices. They have benefited from the efficiency of computer-assisted writing: the ease of revision, the built-in spell checks and grammar checks, and the ease of writing quickly without worrying about legibility. They may have never learned to write in cursive, as this has fallen out of the curriculum in many places. For most of today’s students, writing by hand is more cumbersome, slower, and unnatural. Writing by hand for EAL students may be additionally cumbersome if their home language uses a different writing system.
Second, the time-pressure of in-class writing may also cause additional stress for EAL students, resulting in poorer quality writing (Smith, 2011; Khuder & Harwood, 2015).
Third, it is expected that handwritten work by all students will contain more errors—misspellings, issues with syntax, issues with grammar. Handwritten work by EAL students will also include issues related to language transference—missing or incorrect determiners (a, the), incorrect prepositions, issues with verb tense, issues with word order (Silva, 1993). Together, these kinds of issues may be thought of as “written accent.”
To address these challenges:
- Give students more practice writing by hand. You might create small in-class writing activities, such as tickets-to-enter and tickets-to-leave or exploratory writing tasks. These short informal pieces have the added benefit of allowing you to familiarize yourself with your students’ handwriting before a more high-stakes task, such as an in-class exam.
- Allow EAL students to have additional time to complete in-class handwritten tests and papers, an accommodation recommended by researchers (Solórzano, 2008). Research has shown that students who have additional time don’t have an unfair advantage over other students (Gernsbacker et al., 2020).
- Provide students with practice doing timed in-class writing (Lee & Polio, 2026). This practice will allow students to develop strategies for planning writing, such as jotting down ideas or developing a quick outline before starting to write.
- Read past these written errors and try not to let these minor sentence-level issues impact your assessment of the student’s knowledge. As with all forms of assessment, we need to make sure we are assessing only what we have actually taught (Matsuda, 2006, p. 640). Using a rubric may help keep your focus on your learning outcomes. See this guide by the Ohio State University on responding to EAL student writing.
- Provide students with strategies for preparing for hand-written assessment, such as those found on this ELSO webpage.
Oral assessment
Why use this form of assessment: Since assessing student writing has become more complicated, many instructors are turning to oral assessment (Dobson, 2023). Instructors are either replacing written tests and papers with oral assessment or using oral assessment alongside written assessment, to gauge student’s grasp of concepts they have written about in tests or papers (as a form of an oral defense).
The challenges of using this form of assessment: If using oral assessment alongside written assessment, you might notice a student struggling to discuss what they have written about. This struggle does not necessarily mean that they wrote using AI. Students’ levels of proficiency in writing and speaking may vary widely, depending on how and where they learned English (Reid, 1988). Resident EAL students (students who immigrated to the U.S. during middle or high school) often are stronger speakers of English than they are writers of English. EAL students who are in the U.S. on a visa or more recently immigrated to the U.S. are often stronger writers of English than they are speakers of English.
Further, stress impacts an EAL student’s ability to access their knowledge of the second language. Oral examinations, unlike written examinations, can incur social anxiety (Laurin-Barantke et al., 2016), which may raise a student’s stress level. EAL students may also experience foreign language anxiety—anxiety caused by speaking an additional language in the presence of a native speaker (Stokes, 2025). Time pressure itself may also increase anxiety (as discussed above in relation to timed writing). Anxiety and stress have been shown to impact working memory, causing disfluency (Stokes, 2025).
EAL students’ oral accents may also impact assessment. Pervasive in U.S. culture is linguistic stereotyping—stereotyping based how one speaks. Research has demonstrated that speakers of English as an additional language often face discrimination in the workplace (Hideg et al., 2024) and in higher education (Lourido & Snell, 2025) due to oral accents. Linguistic stereotyping may lead to fewer opportunities for students to speak in class. Native-English speaking students may exclude EAL students during discussions; EAL students themselves may choose not to speak as often in class due to fear of discrimination (Lourido & Snell, 2025, p. 19). EAL students may thus lack experience with speaking in English, leading to a lack of fluency.
It may also be that the instructor cannot understand the student based on their oral accent. Research has shown that a listener’s comprehension is impacted by whether the listener and speaker share a first language (Major et al., 2002) and whether the speaker has learned to focus on sentence stress, pitch, and syllable rhythm, which can greatly impact listeners’ comprehension (Derwing & Munro, 2015). The stress of the assessment situation itself can cause the speaker to lose focus on these elements of clear speech.
What to do about it:
- Help students prepare for oral assessment by providing low-stakes opportunities for practice during class. You may put students in small groups, give speakers within the groups clear roles (so that no student is left out), and provide speaking topics similar to those included in the assessment.
- Keep assessment focused on content, not accent or fluency. Try not to conflate oral accent and fluency with knowledge of the work being assessed. During assessment, keep your focus on the project and course learning objectives. Using a rubric could help with this.
- Work with the student to ensure clear communication. To help the student understand which parts of what they are saying is and is not being communicated, you may repeat the parts of the student’s utterance that you understand and ask for clarification on the parts you did not catch.
- Pair oral assessment with written work. As recommended by Cornell's Center for Teaching Innovation, "Including graded written work, such as an annotated bibliography, research report, and a script at the various milestones, will allow students to show their progress and demonstrate their understanding of the learning objectives while reducing the stress of having the entire grade based on the final product." See CTI's useful guide to oral assessment for more ideas.
- Lower the stress by starting the session with small talk and by providing additional time.
- Recommend resources that will help the student work on clear oral communication. You may refer the student to the English Language Support Office’s resources on pronunciation: the course ELSO 6210 Improving Pronunciation and ELSO’s tutoring service on pronunciation. You could also recommend AI tools and apps, such as those reviewed on this handout.
- Provide students with strategies for preparing for oral assessment, such as those found on this ELSO webpage.
Class Participation
Why use this form of assessment: In the age of AI, a trend is to shift portions of the course grade from homework to class participation. Like homework, class participation can be used to engage students in course material and promote learning. Further, class participation can be used to facilitate the co-construction of knowledge.
The challenges of using this form of assessment: For many EAL students, class participation is challenging. As mentioned above, accent can interfere with class participation, preventing a student from speaking during class and creating barriers to participation in small group discussion. A student’s language proficiency can impact their ability to keep up with the lecture or discussion. It is very challenging to translate, take notes, and prepare questions and comments in English all in time to pose that question or comment before conversation has moved on.
Further, class participation is cultural. Some students may simply not be accustomed to class participation. Students from cultures where the teacher is seen as the possessor of knowledge may just want the teacher to tell them the answers and may not value responses from peers. These students may not understand class discussion as a process of discovery and meaning making. Questions meant to prompt exploration of ideas (e.g. What is art?) may simply be confusing to these students. They may view class discussions as tests and therefore think that if they can’t say something well thought out that they should stay silent.
What to do about it:
- Be transparent about why and how you are assessing for class participation. Tell students why you value class participation, how you define it, and your expectations for it. For example, talk about how class participation is used in your class (e.g. to build knowledge together, to explore concepts) and assure students that there are no stupid questions or comments. Provide criteria you will be using to assess class participation.
- Allow students to use AI to live transcribe and/or translate a class lecture (with the training turned off). AI tools have this capability. Allowing students to use it will help them keep up with the lecture and be more active participants in class discussion. ELSO prepared a letter students can use to make this request. This letter acts as a contract, and asks the student to agree to use the technology ethically.
- Provide students with an outline of the lecture or copy of the slides to facilitate note taking. Having an outline will help students anticipate the contents of the lecture and provide a place to take notes, allowing more of their attention to focus on the lecture and thus giving them a better chance at participating.
- Provide students with necessary vocabulary ahead of the lecture. Unfamiliar vocabulary can throw off a student's ability to follow the lecture and could cause them to look up words during class, breaking their focus from class and deterring active participation. Ahead of class, provide students with a reading, the slides, or other materials that they can review.
- Avoid basing an entire class participation grade on speaking. Class participation has multiple elements. In addition to speaking, class participation includes being prepared for class, being an active listener, taking part in class activities, taking part in group activities, and taking advantage of office hours. Including these elements when assessing a student for class participation takes some of the pressure off of speaking.
- Allow students to use office hours as a proxy for speaking in class. For some students, speaking in class may simply be too challenging. Providing these students with the option of instead taking part in office hours gives them a chance to ask questions and talk through course materials while giving you the chance to gauge their engagement with the class.
- Recommend resources that would help the student build discussion skills. These resources include courses: ELSO 6220 Taking Part in Discussions and ELSO 6230 Designing and Delivering Effective Presentations (which can help the student get more comfortable with public speaking). These resources also include the English Conversation Hours (drop-in sessions to practice social English and discussion), which are co-sponsored by the Language Resource Center and ELSO.
- Provide students with strategies for preparing for class discussion, such as those found on this ELSO webpage.
Computer-assisted writing (e.g. take-home tests, essays, reports):
Why use this form of assessment: Even in the age of AI, there are many good reasons to continue assessing student knowledge through take-home tests, essays, reports, and other types of writing students complete while using a computer. Writing has long been used in U.S. higher education to promote learning and assess student knowledge. And there are many reasons for this. Writing itself is a mode of thinking, allowing us to clarify our ideas, reflect on our ideas, making connections among ideas, and better remember our ideas. Writing rhetorically—in relation to an audience within a specific context—allows us to learn how to communicate more effectively, impelling us to consider our stance, situate our ideas in relation to ongoing conversations about the topic, and consider language choices (INWAC, 2014, p. 5).
The challenges with this form of assessment: In the age of AI, it can be difficult for instructors to discern which content in student’s work originated with the student and which originated from AI (Cotton, Cotton, & Shipway, 2024). And AI detectors don’t help, as research has demonstrated that they often falsely flag student writing as AI generated, especially the writing of EAL students (Liang et al., 2023).
What you can do:
- Create clear AI policies: In alignment with the Center for Teaching Innovation, we advise you to be very clear in your policies about which AI uses are allowed and are not allowed. To create this policy, you might refer this list of ways that AI is being used for writing. You might also refer to other Cornell policies and guidelines:
- Define what you mean if you are allowing students to use AI for editing and grammar: If students upload their papers into AI asking for revisions to grammar, the AI tool will do a full re-write of the paper, adjusting not only grammar, but also style, word choice, and structure. The resulting paper will have an “AI voice” (“Ainglish”), which may lead you to question the student’s role in writing the paper.
- Ask students to use prompts that prevent AI from revising their work. If you would prefer to hear more of the student’s voice, consider asking them to use these prompts:
- Prompt 1: I am going to upload my paper. Do not summarize it or make any comments or suggestions until I provide a prompt.
- Prompt 2: Create a table. In the first column identify grammar errors and typos. In the second column, identify the paragraph number and sentence in which you found the problem. In the third column, explain why this is a problem. Do not provide a correction or rewrite the sentence.
- De-emphasize Standard Written English: If you prefer that students not use AI at all, consider taking the pressure off of students, so that they do not feel they need to hand in perfectly edited writing. Consider adding a statement to the syllabus or the assignment saying that you prefer authentic human writing (with all its flaws) to AI writing.
- Minimize the portion of the grade focused on proof-reading. There may be times that you do want to emphasize editing so that students gain these skills. If so, assign only five percent of the grade to proofreading (using a rubric can help with this). Provide students with tools and to meet this goal, such as the prompts listed above, your own feedback to an early draft, or feedback from a tutor.
- Recommend resources that will help the student work on writing skills. You may refer the student to the English Language Support Office’s resources on writing: ELSO's courses on writing and ELSO's tutoring services on writing and on writing with AI (these tutors can also help students wean themselves off of AI).
- Provide students with strategies for preparing for computer-assisted writing, such as those found on this ELSO webpage.
References
Barkaoui, K., & Knouzi, I. (2018). The effects of writing mode and computer ability on L2 test-takers’ essay characteristics and scores. Assessing Writing, 36, 19–31.
Center for the Study and Teaching of Writing. (n.d.). Responding to international student writing. College of Arts and Sciences. The Ohio State University. https://cstw.osu.edu/responding-international-student-writing
Cotton, D. R. E., Cotton, P. A., & Shipway, J. R. (2024). Chatting and cheating: Ensuring academic integrity in the era of ChatGPT. Innovations in Education and Teaching International, 61(2), 228-239. https://doi.org/10.1080/14703297.2023.2190148
Derwing, T. M. & Munro, M. J. (2015). Pronunciation Fundamentals: Evidence-based Perspectives for L2 Teaching and Research. Netherlands: John Benjamins Publishing Company.
Dobson, S. (2023, April 19). Why universities should return to oral exams in the AI and ChatGPT era. The Conversation. https://theconversation.com/why-universities-should-return-to-oral-exams-in-the-ai-and-chatgpt-era-203429
Gernsbacker, M. A., Soicher, R. N., & Becker-Blease, K. (2020). Four empirically based reasons not to administer time-limited tests. Translational Issues in Psychological Science, 6(2), 175–190. https://doi.org/10.1037/tps0000232
Hideg, I., Shen, W., & Koval, C. Z. (2024). Hear, hear! A review of accent discrimination at work. Current Opinion in Psychology.
Hunter, W. (2026, Sep). The blue book is back. The Atlantic, 338(3). 11-13.
INWAC. (2014). Statement of WAC Principles and Practices. https://wacassociation.org/wp-content/uploads/2019/11/Statement-on-WAC-Principles-and-Practices.pdf
Khuder, B. & Harwood, N. (2015). L2 writing in test and non-test situations: Process and product. Journal of Writing Research, 6(3), 233-278. http://dx.doi.org/10.17239/jowr-2015.06.03.2
Laurin-Barantke, L., Hoyer, J., Fehm, L., & Knappe, S. (2016). Oral but not written test anxiety is related to social anxiety. World journal of psychiatry, 22(6), 351-357. DOI: 10.5498/wjp.v6.i3.351
Lee, H. K. (2004). A comparative study of ESL writers’ performance in a paper-based and a computer-delivered writing test. Assessing Writing, 9(1), 4–26.
Lee, J. & Polio, C. (2026). Effects of varying time constraints on quality, linguistic features and fluency behaviors in L2 writing. Assessing Writing, 70, 1-19. DOI: 10.1016/j.asw.2026.101123
Liang, W., Yuksekgonul, M., Mao, Y., Wu, E., & Zou, J. (2023). GPT detectors are biased against non-native English writers. Patterns, 4(7), 1-4. https://doi.org/10.1016/j.patter.2023.100779
Lourido, G. T. & Snell, J. (2025). Accent bias and inequities in higher education: A critical perspective on university students’ experiences of accent-based disadvantages. Pre-print. https://osf.io/preprints/osf/sze2k_v1.
Major, R. C., Fitzmaurice, S. F., Bunta, F., & Balasubramanian, C. (2002). The effects of nonnative accents on listening comprehension: Implications for ESL assessment. TESOL Quarterly, 36 (2), 173–190. https://doi.org/10.2307/3588339
Matsuda, P. K. (2006). The myth of linguistic homogeneity in U.S. college composition. College English, 68(6), 637-651.
Reid, J. (1998). ‘Eye’ learners and ‘ear’ learners: Identifying the language needs of international students and U.S. resident writers. In J. Reid & P. Byrd (eds.), Grammar in the Composition Classroom (pp 3–17). Heinle and Heinle.
Silva, T. (1993). Toward an understanding of the distinct nature of L2 writing: The ESL research and its implications. TESOL Quarterly, 4(Winter), 657-677. https://doi.org/10.2307/3587400
Smith, C. (2011). Examinations and the ESL student: More evidence of particular disadvantages. Assessment & Evaluation in Higher Education, 36(1), 13–25. https://doi.org/10.1080/02602930903173959
Solórzano, R. W. (2008). High stakes testing: Issues, implications and remedies for English language learners. American Educational Research Association,78(2). https://doi.org/10.3102/0034654308317845
Stokes, J. D. (2025). An exploration of how anxiety and working memory affect international students’ English oral communication. Dissertation. University of Bristol.