scieee AI-readable full text Open interactive document viewer

Data to Discovery Series (Fall 2025)

Li, Xiuqi

Abstract

This collection hosts the slide decks from the UCSB Library’s Data to Discovery Series (Fall 2025). Held every two weeks, these 30 minute sessions highlighted practical skills for working with data and engaging in open scholarship. Each session offers a brief introduction to tools, resources, and good practices — along with space to learn from experts and connect with peers. Fall 2025 lineup: Oct 9 | R or Python? Let’s talk! 🐍📊 Oct 23 | Keys to Your Data: Understanding Ownership, Access, and Use 🔑 Nov 6 | Nightmare on Data Street 🔪 Nov 20 | Your Research Fingerprint: ORCiD, DOI, and more 🆔 Dec 4 | Wrapping Up Your Data Year 🎁 Slides were used for educational purposes. We invite other educators, librarians, and community facilitators to share, reuse, and adapt these materials for training and outreach.

Full text

Nightmare on Data Street 👻 #Halloween Research Data Services [email protected] Nov 6, 2025 Data to Discovery | Fall 2025 🎃 🎃 �� RIP What we are not talking about today… Research Misconduct Screenshot from ORI, HHS NSF uses a similar definition Questionable Research Practices Prevalence of self-reported QRPs (Gopalakrishna et al. PLOS One 2022) QRP1: Insufficient attention to equipment, skills or expertise. QRP7: Inadequate note taking of the research process. QRP10: Insufficient inclusion of study flaws and limitations in publications What’s in Store on Data Street 🕯 💬 Story: We’ll look at a short “data nightmare” scenario 🤔 Solution: You’ll use the Zoom annotation tool to vote on possible fixes (more than one may apply) 🛠 Resources: We’ll highlight a UCSB service that can help prevent similar problems in your own work Sharing Spell Gone Wrong The lab keeps a shared Box folder with subfolders for each lab member. One day, a PhD student went to retrieve some data and analysis from an early project to prepare a conference poster — only to find that the results didn’t make sense. After some detective work, the PhD student discovered that another lab member was reusing the same analysis workflow and had (accidentally) overwritten the original data with their own. The PhD student tried to recover the previous version through Box’s version history… but the file’s history didn’t go back far enough… Image from Flaticon What could have prevented this? ●Store raw data in read-only or write-protected folders so it can’t be modified easily ●Remind everyone to be careful when saving ●Rename file as FINAL_FINAL_DO_NOT_TOUCH ●Follow the 3-2-1 backup rule: keep 3 copies of the data, on 2 different media types, with 1 copy stored offsite ●Automate routine backups or snapshots of shared storage ●Move everything into one master folder for easier access CrashPlan An automatic backup service that continuously protects your data by securely backing up files from your computer to the cloud. **Free to UCSB researchers For more info, see Data Literacy Series on CrashPlan and GRIT CrashPlan FAQ Protocol of Shadows Slide adapted from Reproducibility for Everyone (Shared under CC-BY license) What would help avoid “protocol of shadows”? ●Rely on on the senior grad student — they know how it’s done! ●Use a shared electronic lab notebook (ELN) or collaborative documentation platform with version tracking, so protocol updates are visible to the whole team ●Once a new protocol is approved, delete all earlier versions to avoid confusion ●Contact the original author ●Email the latest protocol to everyone on the team so they can save their own copy ●Publish detailed protocols or methods in a dedicated repository Protocols.io An open-access, collaborative platform for creating, managing, and sharing step-by-step research methods/protocols in a dynamic, interactive format. **UCSB researchers have access to the premium plan See Getting Started with Protocols.io at UCSB for instructions to set up an account and access the UCSB premium license Screenshot of Lisa Mesrop. protocols.io. 2024.