Nupoor Ranade
George Mason University
This assignment asks students to generate a complex essay using an AI text generation tool, edit the essay using principles taught in class to improve the readability score of the generated content. Students are asked to share the final output along with visuals that demonstrate the comparison between the various versions of the generated content. This assignment can be adapted for all course levels, especially for first-year writing and professional and technical writing classrooms.
Learning Goals:
Original Assignment Context: cross-listed graduate and undergraduate class on technical editing
Materials Needed: MS Word, screen capturing tools, and any one online tool to generate content from the following list:
Time Frame: ~1-2 weeks
Introduction
This chapter is based on a pedagogical experiment using AI writing tools conducted in a cross-listed graduate and undergraduate class on Technical Editing in George Mason University’s Professional and Technical Writing program during Fall 2023. It discusses an assignment that required students to: 1) generate a short essay (minimum 300 words) using AI software, 2) edit it manually using editing principles taught in class (such as clarity, conciseness, accuracy of grammar and punctuation, and sentence formation) and 3) improve the readability score of the essay’s content by reducing the the Flesch-Kincaid Grade Level. This metric is equivalent to the US grade level of education and shows the required education to be able to understand a text. Students harvested the required content using an installation of AI tools like GPT-2, and performed the three tasks. Their submission displayed markup that recorded their work and the screenshots of readability scores from before and after they made edits. Students also provided justifications on the edits made with considerations of the writer being an AI tool. This assignment made students think about AI editing beyond simply analyzing automatic editing features of AI tools like Grammarly.
The topic of using artificial intelligence (AI) tools for professional writing courses resulted in an informal debate in the Technical Editing class that I taught at George Mason University in Fall 2022. There were two contrasting opinions: the first group (with fewer members) argued that the job of a corporate editor is to some extent disappearing owing to changing technologies. Kreth and Bowen’s (2017) finding that 75% of the participants perform copyediting and proofreading tasks using macros (computer programs) supports this argument. This number was up by 40% from a similar survey in 1999 exhibiting an upward trend when fewer editors (only 35%) used software programs. Group 2, comprising majority students in the class, contended that AI editing tools can never fully replace human editors. Drawing from their readings in the class, they argued that human editors work closely with writing teams to make sure that the content aligns with writers' goals and audience expectations thus “linking the author and the potential reader, and serving the needs of both” (Buehler, 2003 p. 463). Since AI lacks these insights, it would be challenging for it to replicate this work. This assignment was created as a response to the debate, and to create meaningful discussions about “human-in-the-loop” in AI technologies, and to review the limitations of tools used in technical and professional communication settings.
AI writing is still in its early stages and far from perfect; yet it is able to produce texts that are indistinguishable from that of a human writer. Until now, electronic editors or software editing programs/applications were studied only to highlight their capabilities such as speed, copyediting functions, version control, and features that afford human collaboration. With AI, technical editing pedagogy needs to reflect the new realities of the use of tools that are used for controlling, managing (Flanagan & Albers, 2019) and creating texts.
AI word processors can automatically edit spelling and grammar errors as you type. Natural Language Processing (NLP) and rule-based engines allow such tools to help users identify errors in language (grammar and sentence structures) and mechanics (such as punctuation, capitalization, abbreviations) or fix them automatically. Grammarly and Quillbot are popular examples that use this technology. These tools also have plugins that support browsers or word processing software (Fitria, 2021). These features significantly speed up the copyediting process and cut back the number of revisions required for written drafts. The use of AI tools in research writing helps deal with some humanistic concerns in writing, such as motivation and anxiety. While it is possible to interpret a human editor’s feedback in a negative way, especially if it does not reflect components of effective dialog like empathic understanding of writer’s goals and unconditional positive regard (Masse, 1985), an AI tool’s automated feedback is almost always perceived as unbiased or indifferent (Rudenko-Morgun, Arkhangelskaya & Makarova, 2023). These features of AI tools are becoming more effective and accurate over time since NLP algorithms are corpus-based; the size of databases is growing with each instance of use, thereby improving the training data. Technological advancements are giving rise to new and more advanced features as well. Tools like Acrolinx used by corporations go beyond language and mechanics. They enable users to render textual content more findable, readable and consistent by checking it against a predefined set of style rules (style guide). In most cases, editors train such tools to reflect the corporation's style preferences. This is the most visible process that exhibits the role of human editors in AI editing. It is true that editors use technologies convenient to either the employer, the writer, or both, and in some cases, corporations may mandate the use of technology that is not preferred by the writers or reviewers, and so may have to find ways to supplement it (Lanier, 2019). Therefore, editors must understand the role of technology as well as their function as humans-in-the-loop to ensure the best interests of writers and their audience.
As AI writing moved beyond professional spaces (like auto-compose in Gmail) to essay writing in classrooms, scholars started making explorations in the field of AI literacy. The discussion of writing with AI tools, or assistive writing, has been a popular line of study in literacy which dates as far back as 2007 (Sternberg, Kaplan, & Bork, 2007). Beyond composing, literacy scholars have studied AI technologies by analyzing algorithmic design and big data perspectives to understand students’ experiences of reading and writing with algorithms, especially with respect to identity, agency, authority, adversary, conversational resource, audience, and so on (Leander & Burriss, 2020). To study human interactions with AI in education we need more rigorous engagement with changing technologies, as well as new ways of conceiving digital literacies than are found in representational paradigms (Leander & Burriss, 2020). To do so, scholars rely on approaches such as Actor Network Theory, posthumanism, assemblages, etc. inspired by new materialism and other media theories. Although such approaches open possibilities to survey different heterogeneous elements in these socio-technical systems, they can be overwhelming and well beyond scope for classes focused on praxis (such as this Theory and Practice of Editing class).
However, an approach like writing and editing with AI can force students into a heightened awareness of our dependencies on technology,posthuman dependencies, that not only ask us to reexamine our definitions of writer, text, and reader, but also to reevaluate our very identities within technological systems (Fyfe, 2022). This assignment helps engage students with the following questions:
Assignment goals
This teaching experiment invites students into an urgent conversation about the role of AI in their professional lives. While pursuing hands-on courses such as technical editing, students find themselves right in the middle of a relationship between entities who attribute agency to each other – writers, editors and audiences. They see the relationship impacted by and/or drastically challenged by AI (Miller 2007). The primary goal of this assignment is to help students recognize their role in such situations, and actively participate in the socio-technical relationships by answering the following questions:
This assignment was not meant to settle a debate on “whether AI will eliminate editor roles from organizations,” instead it focused on immersing students within the debate to broaden and deepen their perspectives about partnering with AI and reflecting on their own intellectual contributions in such roles.
Assignment Requirements: Software and Tools
Deliverables: 1 edited document (essay.DOCX), 2 Screenshots (before.png, after.png)
Deadline: 11:55 PM on <mm/dd/yyyy>
Submission Location: Upload Blackboard
Overview
The most important aspect of technical editing is learning how to evaluate a text at a level higher than sentence level. In other words, how to perform a comprehensive edit. Learning to perform comprehensive editing is learning how to evaluate a text's structure and analyze the ability of that structure to effectively communicate with the document's audiences within their context. This assignment helps you draw from the different techniques and strategies you have learned in class and apply them to content generated by an AI tool, to realize your editing potential beyond sentence-level editing which is, in most cases, handled by the AI text generator itself.
What Am I Supposed to Learn Through This Assignment?
This assignment gives you the opportunity to analyze the opportunities and challenges of AI technologies in the field of professional and technical writing situations. It helps you utilize your knowledge of the various elements of the rhetorical situation such as the genre, audience, writer, purpose, and context. This holistic understanding of what aspects of a content development process generate a fitting response can help you use AI tools as an extension of your capabilities as an editor.
Steps to Complete the Assignment
Generate a 300-word essay on any topic using any one of the free AI tools discussed in class for a public audience. Copy the content in a word document and conduct a readability score check using MS Word and take a screenshot of the statistics. Make edits for clarity, conciseness, and grammar appropriateness until you have reduced the Flesch-Kincaid Grade level identified in the previous step by at least 1 grade. For example, if the initial grade level was 14.4, it must be lowered to a 13.9 or lower. Use Track Changes to make these edits so that the changes are recorded. Take screenshots of readability score as you make edits to ensure the values are decreasing and not increasing as you make edits.
The Deliverable
The final submission will include three files:
Evaluation
The general expectations for editing samples can be found on the course syllabus, including what level of work typically is associated with grades at the A, B, and C range. In evaluating this assignment, I will specifically look for the following issues that correspond with the assignment requirements and our learning objectives in the course:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
References
Albers, M. J., & Flanagan, S. (2019). Editing in the modern classroom: An overview. Editing in the modern classroom, 1-14.
Beuhler, M. F. (2003). Situational editing: A rhetorical approach for the technical editor. Technical communication, 50(4), 458-464.
Fitria, T. N. (2021). Grammarly as AI-powered English writing assistant: Students' alternative for writing English. Metathesis: Journal of English Language, Literature, and Teaching, 5(1), 65–78.
Fyfe, P. (2022). How to cheat on your final paper: Assigning AI for student writing. AI & SOCIETY, 1-11.
Kreth, M. L., & Bowen, E. (2017). A descriptive survey of technical editors. IEEE Transactions on Professional Communication, 60(3), 238-255.
Lanier, C. R. (2019). Concepts in Technical Editing Technologies: What's Important in Practice? In Editing in the modern classroom (pp. 128-145). Routledge.
Leander, K. M., & Burriss, S. K. (2020). Critical literacy for a posthuman world: When people read, and become, with machines. British Journal of Educational Technology, 51(4), 1262-1276.
Masse, R. E. (1985). Theory and practice of editing processes in technical communication. IEEE transactions on professional communication, (1), 34-42.
Miller, C. R. (2007). What can automation tell us about agency? Rhetoric Society Quarterly, 37(2), 137-157.
Sternberg, B. J., Kaplan, K. A., & Borck, J. E. (2007). Enhancing adolescent literacy achievement through integration of technology in the classroom. Reading Research Quarterly, 42(3), 416–420.