Teaching Research in the Age of AI
Teaching Research in the Age of AI
Megan Heuer’s strategies for research assignments that reward process, build skills, and create accountability, plus four ready-to-use student assignments.
Leila Elamine
Content & Faculty Advisory Board Lead, Consensus
Research assignments ask students to enter an unfamiliar field, decide which sources matter, understand complex arguments, and develop a position of their own. Generative AI makes it easy to skip that work. In our webinar Designing Effective Assignments in the Age of AI, Megan Heuer, Director of Educational Initiatives at SMU Libraries, shared how she designs research assignments that students are more likely to do themselves.
We turned her session into a faculty guide and four student assignments you can use as they are or adapt for your course.
Download the toolkit at the top of this page. It includes the faculty guide, four student assignments, two Excel graphic organizers, and Megan’s research paper assignment and rubric.
What’s in the Toolkit
Faculty guide. Megan’s four strategies, her Consensus prompts, and how to put the assignments together.
Map a Research Topic. Students ask Consensus for the major areas of research on a topic, draw a concept map by hand, and choose a focused question.
Make Sense of a Research Article. Students diagram an abstract, follow a strategic reading path, and annotate what matters.
Track Evidence in a Research Article. Students record four pieces of evidence from one article, with page numbers, in their own words. Includes an Excel graphic organizer.
Organize and Connect Research Sources. Students compare three sources in a synthesis matrix and describe the pattern across them. Includes an Excel graphic organizer.
Companion resource. The research paper assignment and final-paper rubric by Megan Heuer and Julia Thoummakat, Social Science Research Librarian, SMU Libraries.
Each assignment comes as a PDF to hand out and a Word version students can fill in.
Why Students Outsource Research
Many students weren’t doing this work before generative AI, either. The Citation Project studied how undergraduates use sources in their research papers. In 94% of citations, students quoted, copied, or paraphrased a single sentence instead of summarizing the source’s argument, and 69% of citations came from the source’s first two pages.
Source: Sandra Jamieson and Rebecca Moore Howard, “Sentence-Mining: Uncovering the Amount of Reading and Reading Comprehension in College Writers’ Researched Writing,” in The New Digital Scholar, ed. Randall McClure and James P. Purdy (Information Today, 2013).
The same figures are reported in Sandra Jamieson, “Reading and Engaging Sources: What Students’ Use of Sources Reveals About Advanced Reading Skills,” Across the Disciplines 10, no. 4 (2013).
Megan opened the webinar by asking attendees why students outsource research assignments to AI. Their answers included time pressure, unfamiliar topics, low confidence, worry about grades, and doubts that the assignment was worth doing.
“There is no such thing as a bulletproof assignment.”
— Megan Heuer
Instead, Megan combines four strategies.
1. Reward Process and Growth
Students will misunderstand papers, choose irrelevant sources, and follow questions that go nowhere. Megan expects this and builds her course around it. Research is also emotional. Carol Kuhlthau’s model of the information search process shows that students feel uncertain when they begin, become confused or frustrated while they explore, and reach clarity only once they narrow their focus.
Figure 1. Kuhlthau’s model of the information search process. Original model by Carol C. Kuhlthau.
Megan tells her students that these feelings are a normal stage of research, not a sign they’re doing it wrong. She also stages her research assignment so early attempts are low stakes:
Research question
Synthesis matrix
Outline and annotated sources, followed by a group consultation
Draft
Final paper
The early stages are marked excellent, sufficient, needs work, or missed, with feedback. Only the final paper counts toward the grade, but students must complete each stage before moving on. Her final-paper rubric includes a process criterion that scores how well students used feedback and how much they grew.
2. Gain Student Buy-In
Many students don’t know what “critical thinking” means in practice. Megan recommends naming the specific skills an assignment builds, the ones students will use in their careers: evaluating conflicting evidence, recognizing uncertainty, asking more precise questions, and revising a position when the evidence changes.
Choice helps too. Students might choose their topic, research question, audience, or final format. The final product doesn’t have to be a paper. Zines and online exhibits still require students to find, evaluate, interpret, and synthesize evidence.
3. Build the Skills Students Need
Sometimes outsourcing signals a missing skill. Megan sees reading as one of the biggest barriers to good student research, and she teaches it through modeling and practice. She starts with a single abstract and shows students where the findings are, since abstracts often open with what other researchers have done.
“Where do your eyeballs go next?”
— Megan Heuer
She asks this while modeling how to scan a paper or a list of search results, so students learn which parts to read closely and which to skim. Academic librarians specialize in teaching research skills, and Megan encouraged faculty to reach out to theirs.
4. Create Accountability
Megan describes accountability as creating friction: making it slower and harder to outsource the parts of an assignment that matter most.
Source links and page numbers in every citation
A stated consequence for fake citations. In Megan’s course, one fake citation means a zero on the paper.
Annotated copies of sources
Research questions, matrices, outlines, and drafts submitted along the way
Consultations, oral presentations, and exam questions about each student’s own research
Each checkpoint takes time to review, so pick the ones tied to your most important learning goals and manageable at your class size.
5. Use Consensus to Show How Scholarship Works
Students often treat research as a set of fixed answers. Megan uses Consensus to show them that researchers disagree, revise, and build on each other. She gives students a prompt, has them run it, then asks what they found and what they learned.
What are the major areas of research on [topic]?
What research disagrees with [claim]?
How has research on [topic] changed over time?
Where are the findings mixed or uncertain?
For a question about how research has changed, Consensus can show a timeline of the papers it found.
Figure 2. A Consensus answer to “How has research on retrieval practice evolved over time?” with a timeline that places each cited paper by its publication year.
Students still need to open each source, read it, and record page-level evidence. The student assignments are built around that step.
Where to Start
The four assignments can serve as the early stages of a staged research project like Megan’s. If you do nothing else, start with the three changes she recommends:
Require page numbers and source links in every citation.
Have students turn in their sources on a graphic organizer.
Have students turn in annotated sources.
“Keep iterating, keep trying. This is important work. Don’t give up on it.”
— Megan Heuer
Want to hear more from Megan? Watch the full webinar.
Become a Consensus MCP expert.
For courses and more information on how to use the MCP, check out our guide below.
