You have determined how long your paper is...
Now that you have determine how long it takes to read your paper aloud, it is time to make some edits.
While there is some variation, a graduate research presentation at a conference or symposium is typically around twenty (20) minutes. At the average rate of speaking, twenty minutes correlates to approximately seven to ten typed pages, when typed in Times New Roman, 12 pt font, double-spaced. In terms of individual pages with the same typeset, one page typically takes the average person about two minutes to read.
Of course, some people have faster or slower speaking cadences, so this may vary. But it is useful baseline to get us started in adapting your paper.
Now that you have a rough idea how long your presentation script should be based on the conference or symposium parameters and your starting paper length, it is time to decide how to cut down your paper.
There are two main types of research presentations: wide overview and specific case study. A wide overview presentation looks at the whole argument from your original paper, excluding specific examples or case studies in favor of presenting the argument. In opposition to this, a specific case study narrows the focus of your argument to center around a strong example, forgoing more minute points in favor of presenting a strong argument.
The type of research presentation you decide will depend on your paper, your preferences for your research, and the specifications of your conference or symposium. Some conferences/symposia will be centered on a particular theme or topic, which may sway your decision one way or another. Events focused on methodology may warrant a specific case study presentation. Likewise, if you are developing your own methodology, a wide overview may be more appropriate.
Who is the audience of your presentation? Are you presenting at a subject-specific conference or symposium where everyone in the audience is a member of a single field or academic discipline? In this case, you can include some subject-specific terminology with the assumption that your audience has a baseline understanding of your topic's background. For example, if you are presenting a paper on new alcohol research in teens at a conference on alcohol research, you can assume your audience is familiar with scientific methods of study and collection in that area.
Or are you presenting at a more general and/or interdisciplinary conference or symposium? In this case, your audience may come from a variety of academic backgrounds so it is best to assume they know nothing about your topic when editing your paper. Subject-specific terminology should be defined and you should prepare brief descriptions for any concepts that may be unknown. When in doubt, always over-explain. For example, if you are presenting the same paper from above on new alcohol research in teens at a conference displaying graduate research across all disciplines, you cannot assume that your audience is going to know what gene marking is.
Where is your presentation taking place? Will you be in a lecture hall where you can provide images and other visual material for your audience to look at? Or will you be in a conference room with no projector? Will you have a microphone? Is the event a circulating one where you should expect audience members to come in and out of the room during your presentation, or will there only be a single presentation at a time? If there are multiple presentations running concurrently, how will you keep your audience's attention? Some of this information you may have to get from the organizer of the conference or symposium you will be attending.
Using the template (via Google Docs*) provided below, go through your paper paragraph-by-paragraph and identify what material is integral to your argument and what is supporting examples or details. Paragraphs that are key to building and supporting your argument are sorted into the left column; paragraphs that are extra support are sorted into the right column. The paragraphs that you sort into the column on the right are meant to be edited down to save space or deleted entirely from the final presentation.
Remember: at this point you are not actually editing any of the paragraphs. Keep the text the same and focus on what each paragraph is saying rather than its grammar or flow. Feel free to be brutal; you may find yourself needing to cut entire sections.
*Note: make a copy of the document in Googles Docs to edit it yourself.
Alternatively, this portion of the module can be completed using Generative AI either exclusively or in tandem with completing the activity on your own as a second opinion. The following prompt can be pasted into any LLM of your choice, with the edits suggested:
"Read the paper I have attached and suggest portions to cut from the paper to make it [INSERT NUMBER] pages long while keeping the core argument. Go through the paper paragraph-by-paragraph. [PICK ONE: Focus on one example to support the core argument. (--for specific case study papers) // Focus on the paragraphs that provide an overview of the paper's argument with brief supporting examples. (--for wide overview papers)] Do not add any new text to the paper. Do not edit any of the text already in the paper. Only use the text and words already in the paper. Present the suggestions of which paragraphs to keep or edit/delete as a table with two columns: paragraphs to keep on the right and paragraphs to edit/delete on the left. Provide a brief justification for keeping or editing/deleting each paragraph in the table."
Where text is in brackets '[]', fill in the text. This prompt is designed to produce a completed 'Paragraph Splitter' assignment table, though some variation may occur. The prompt is also designed to not add or change any text in your paper, only identify potential paragraphs to keep or edit/delete in your presentation. Be sure to upload your original paper in the prompt. You do not need to include the figures or images.
Any LLM platform can be used to complete this activity, but I highly recommend a platform with data privacy to ensure your research is not scrapped. Certain tiers of LLM platforms such as ChatGPT or Claude include these privacy protections. Additionally, UNC has its own LLM 'sandbox' called PromptLab that allows UNC-affliated users to work with various LLMs within a data-secure space.
Below is an uploaded and completed version of the 'Paragraph Splitter' activity for the example paper. In the case of this paper, it is best to provide substantial cuts in the visual analysis section (the final presentation will be accompanied by images) and an additional argument that is not essential to the main argument. Smaller edits for brevity will also be made.