The purpose of the course is to make students better scholars to successfully complete a M.S. or PhD thesis. For PhD students, the course is a preparation for their Research Proficiency Exam (RPE).

You don’t just learn about scholarly work in this sequence, you produce it. In the fall, you build a research prospectus for your thesis. In the winter, you turn one part of that plan into a conference-style research paper and a research talk.

Doing research together with AI is the focus of the redesigned sequence. Every module has you work with AI tools on that week’s real task and then check what they gave you. You will search the literature with them and verify what they find, interrogate papers you have already read, have your research questions attacked by a skeptical critic, stress-test your evaluation design, write and check analysis and plotting code, draft and revise sections, and run a review of your own paper before anyone else sees it. You will also learn where they fail: invented references, confident agreement with whatever you propose, and judgments about novelty that cannot be trusted.

What you will produce:

• Fall (ECE 563): a research prospectus for your thesis (about 10-15 pages), built one section per week: your answers to the Heilmeier questions, a documented literature search and a curated library, a literature review that identifies a research gap, a research plan with an evaluation design and milestones, a focus statement, a one-page and a long thesis outline, and a complete IEEE bibliography.
• Winter (ECE 564): a research paper and a research talk. You narrow your thesis plan to one paper, write it in a real venue’s template (4-8 pages), have it reviewed by an AI reviewer, revise it, write a response letter, and record a 10-12 minute conference-style talk.

For PhD students, the paper and talk are direct preparation for the RPE. For M.S. thesis students, the paper is typically a chapter of the thesis.

Doing research with AI:

Learning to work with generative AI is part of this course, not a workaround. The assignments are designed around one principle: AI may assist at every stage, but you own the research.

Where you will work with AI, week by week:

• Finding literature: semantic search and discovery tools, then verifying every reference the tool hands you against the actual paper.
• Reading: your own notes first, then an AI summary, then a comparison that shows you what the tool flattened or got wrong.
• Framing the problem: having a model attack your Heilmeier answers as a skeptical program manager would, then revising.
• Designing the research: asking what would make your evaluation unconvincing to a reviewer, and fixing it before you run anything.
• Code and analysis: generating experiment and plotting code, then checking that it does what you think, with tests you write.
• Figures: drafting them in code, and checking scales, aggregation, and dropped points yourself.
• Writing: drafting sections from your own notes, outline, data, and sources, then editing for argument, evidence, and voice.
• Review: an AI review of your own paper, and deciding which comments to act on.

Four rules make it work:

• AI works from your material. Your notes, outline, data, code, and the sources you have read are the input. Anything an AI tool adds on its own is unverified until you check it.
• You own every claim and every sentence. You verify each reference, number, and result, and you must be able to explain anything you submit.
• You show your process. Every submission includes a short AI-use log, and drafting assignments include your tracked edits from the AI draft to your final text.
• You disclose the use of AI as your publication venue requires.

Using AI is your choice. Nothing in this course requires you to let an AI tool write or analyze for you, and you can earn full marks without doing so. What the course does ask is that you understand these tools: the assignments that have you check, compare, or critique AI output are part of the work, and a free account is enough for them. A paid subscription to a state-of-the-art model is recommended but not required.

Prerequisites/Co-requisites:

• You must be proficient in written and spoken English before taking this course. The course teaches scholarly writing, not English as a second language.
• You must be a PhD student or an M.S. thesis student.
• You must have a thesis advisor by the time the course starts. Several assignments ask for a short confirmation or feedback from your advisor.
• You must have completed some initial research for your thesis before you take this course sequence. If you don’t have results yet, you can write your winter paper about planned work or reproduce a result from a published paper.
• You need access to at least one generative AI tool (free options are fine).
• The course sequence is a requirement for both ECE PhD and M.S. thesis students.

Description:

Scholarly Skills I focuses on planning your thesis research; Scholarly Skills II focuses on writing it up. Working with AI runs through both terms, at every stage of the research process, and so does the verification that has to go with it. Topics covered include research tools, research ethics and integrity, including the responsible use of AI, choosing and planning a research project, working with your advisor, literature search and review, research design and reproducibility, citing sources and bibliographies, thesis structure and outlining, drafting and revising with AI, figures and visualizations, the sections of a research paper, peer and AI review, revision, and delivering a research presentation. At the completion of the sequence, students are familiar with the tasks and activities for successfully completing a M.S. or PhD thesis, and with what AI can and cannot do for each of them.