Technology Supports for Scholars

• I use these tools most days for writing support, data work, coding, website development, and course design
• I study how to best use technology, and AI in particular, to help people learn
• My default is curious and cautious
• Don't mistake what I say for department, college, or university policy

• Every section is a tile on the map
• Each slide has its own link. Bookmark what you want
• The page stays up after today at kendallhartley.com/presentations

• 16 of you completed the survey
• Most of you are in coursework, a few are preparing for comprehensive exams

• I am not sure what my field thinks about AI
• I worry I will use it the wrong way and get in trouble
• I want to know the appropriate ways to use it
• I have concerns about academic integrity
• The water and energy cost makes me hesitant
• Sometimes it says it did the work when it did not
• I annotate articles but have no good system to find that knowledge later
• I would like to connect Zotero with Obsidian

• Sign in so My Library, alerts, Scholar Labs, and Quick Read work
• Settings, Library links, add UNLV Libraries. Full-text links appear next to results
• Settings, Bibliography manager, choose BibTeX or RefMan for one-click export to Zotero or another manager
• Zotero Chrome plugin is invaluable

• Advanced Search option
• Left sidebar for a date range (Since 2022) and Review articles
• Cited by moves forward to newer work that builds on a paper. Related articles finds close neighbors
• All versions often turns up a free author copy

1. Menu (top left), then Metrics
2. Top publications ranks journals and conferences by h5-index
3. Select View all
4. Choose a category and subcategory, such as Social Sciences, then Educational Technology
5. Search a journal by name to see where it sits

• Higher means more cited, not necessarily better
• Compare journals within a field, never across fields

1. Click a journal's h5-index to see the articles that make it up, most cited first
2. This is a quick way to find the papers everyone in a subfield is citing
3. Cited by counts on any result give the same signal for a single article
4. Sort what you find into must-read and skim

• Author profiles show total citations, h-index, and i10-index, all time and for recent years
• Follow an author to get new articles and new citations by email
• Create alert on a search or on a key paper's citations
• Start your own profile now

• Scholar counts citations broadly (preprints, theses, slides), so its numbers run higher than Scopus or Web of Science
• Citation counts measure attention, not quality
• Ask your chair which journals matter in your area before using any ranking to choose where to submit

• Scholar Labs takes a full research question in plain language, not keywords
• It breaks the question into parts, searches Scholar, and returns papers with a short note on how each one answers the question
• Quick Read (new in 2026) gives a structured summary of an article to help decide whether to read it
• Both require signing in

• It returns a limited set of papers and does not support Boolean operators
• Good for finding a starting set, not for a systematic review
• A summary is not a reading. Open the article before you cite it

• Reading takes the most time
• Remembering what you read is the second problem
• Notes end up spread across PDFs, folders, and apps

• Chats, folders, notebooks, and projects
• A notebook built on sources you choose (PDFs, Google Docs, web pages, videos)
• Answers come with citations that point to the passage in your source
• Today's demo is NotebookLM (Google now also calls it Gemini Notebook)
• Similar options at UNLV include Claude projects (free to UNLV students through Rebel Spark AI) and Gemini with uploaded files

• One notebook per course, project, or comps question
• Ask questions across all sources at once and click each citation to check it
• Generate a study guide, briefing doc, or mind map from your sources
• Audio overviews turn a set of readings into a podcast-style conversation
• Free accounts allow up to 50 sources per notebook, each up to 500,000 words
• Sign in with your UNLV Google account to possibly activate more sources

• What methods did these studies use, and what were the sample sizes?
• Where do these authors disagree?
• Which of these sources define construct, and how?
• What does source 3 say about topic? Quote it.
• Build a table comparing purpose, sample, and findings

• Zotero holds the articles and your annotations
• A notes app (Obsidian, Google Docs, or Notion) holds your own synthesis
• A grounded AI notebook helps you search across both
• Write Say a one-paragraph summary yourself (Wispr Flow, or use my referral link, which gets me a free month)

• It can miss a key point or blend two sources
• It cannot tell you a study is weak
• A summary is not a substitute for reading the articles you will cite
• Use it to decide what to read closely and to find what you already read

• Your syllabus and your instructor
• Your committee and your chair
• The journal or conference you submit to (most now have AI policies)
• UNLV generative AI guidance for students and the Student Academic Integrity Policy (links in 6.7)
• This is a fluid situation. Guidance tomorrow will likely look different than guidance today

• Note which tool you used and for what
• Keep your prompts and drafts for work you submit
• Prepare and plan for a disclosure statement that conforms to your publication outlet. Example: Hartley, Sáez-Delgado, and Mella-Norambuena, Future Internet, 2026

"During the preparation of this study, the authors used Claude 4.6 for the purposes of developing GenAITutor.net, which was the primary data collection and analysis system. Claude 4.8 was also used to critique methods, analysis, and conclusions and contributed to the final editing process. The authors have reviewed and edited the output and take full responsibility for the content of this publication."

• Verify every citation. Look up the article yourself
• Check that a quote actually appears in the source
• If it says it read, ran, or saved something, confirm it did
• Ask it to show its work and point to the source
• Reviewers will punish hallucinations mercilessly

• Do not upload participant data, identifiable information, or student records to consumer AI tools
• Your IRB protocol governs where research data can go
• UNLV's rule is plain. Never enter FERPA or HIPAA protected data, or personal and sensitive information, into any AI tool, even one covered by a UNLV agreement
• UNLV-supported tools (Claude, Gemini through Rebelmail, ChatGPT, Google AI Pro) keep chats out of model training and away from human review. That is better protection, not permission for sensitive data
• A Freedom of Information Act request may make your prompts front page news

• AI use has real energy and water costs (SemiAnalysis, From Tokens to Burgers)
• Reasonable people weigh this differently
• Choosing not to use AI is a legitimate choice for a scholar

• Only 4 of 16 of you use one
• Zotero is free, works with Word and Google Docs, and grabs citations from the browser
• Pair it with Google Scholar search and the Zotero Chrome plugin
• Zotero connects to Obsidian through community plugins

• This presentation, kendallhartley.com/presentations
• UNLV generative AI guidance for students
• UNLV Student Academic Integrity Policy
• UNLV AI tools and data protection
• UNLV AI best practices
• UNLV Libraries AI 101 guide
• UNLV Libraries Ask Us
• UNLV Libraries dissertation and thesis support
• UNLV Libraries Citation Manager Showdown (Zotero and others)
• UNLV Writing Center AI statement
• Google Scholar search help
• Google Scholar Metrics help
• NotebookLM help
• Zotero documentation