Lightnews — Scholar-powered news

Crystal Lewis

@cghlewis.bsky.social

5.5K followers 1.7K following 1.3K posts

Research Data Management Consultant | cghlewis.com Co-organizer @r-ladies-stl.bsky.social‬ Co-organizer POWER Data Management Hub | https://osf.io/ap3tk/ Author of DMLSER: https://datamgmtinedresearch.com/ RDM Weekly: https://rdmweekly.substack.com/

osf.io

Posts Media Videos Starter Packs

Pinned

Crystal Lewis @cghlewis.bsky.social · Oct 21

Re-introduction for new followers!
Hello! 👋
I am currently a freelance research data management consultant. I also co-organize R-Ladies St. Louis. I mostly post about data management and #rstats data wrangling tips. I also recently wrote this book.
datamgmtinedresearch.com

Welcome | Data Management in Large-Scale Education Research

This is the in-progress version of Data Management in Large-Scale Education Research.

datamgmtinedresearch.com

4 30 110

Reposted by Crystal Lewis

Simon P. Couch @simonpcouch.com · 30m

ICYMI, @sara-altman.bsky.social and I have been writing a biweekly newsletter on AI and open source data science on the @posit.co blog!

A bit about how that came to be on my #rstats blog: www.simonpcouch.com/blog/2025-10...

2 8

Reposted by Crystal Lewis

Research Culture Leeds @researchcultureuol.bsky.social · 3h

Day 2 of ##Hiddenref2025 and a reminder that people are as important as outputs.

1 5 7

Reposted by Crystal Lewis

Eli Pousson @elipousson.bsky.social · 12h

Before I code something from scratch, does anyone have a #rstats function like setdiff but it works with named lists and/or data frame rows? Optionally dropping duplicate values and keeping the differences from the first list?

3 2 3

Reposted by Crystal Lewis

Noam Ross @noamross.net · 18h

A month or two ago someone posted a link to their really amazing set of LLM system instructions for writing #rstats code with good tidy/NSE patterns. (They were also good for humans!) Does anyone recall who or where that was?

2 6 15

Crystal Lewis @cghlewis.bsky.social · 18h

No one prepares you that when you work for yourself, you no longer have IT support available to you. But today, thanks to YouTube, I became my own IT person. 😂🙏

4 27

Crystal Lewis @cghlewis.bsky.social · 18h

a man is asking a woman if she is crying or something .

ALT: a man is asking a woman if she is crying or something .

media.tenor.com

Crystal Lewis @cghlewis.bsky.social · 19h

All blessing, no curse. :)

2 1

Reposted by Crystal Lewis

Kristin Briney @kbriney.bsky.social · 1d

Today on my blog: some thoughts on the best data management strategies for collaboration: dataabinitio.com?p=1204

What's your best data tip for collaborative research?

Data Management for Collaborations » Data Ab Initio

dataabinitio.com

3 7

Crystal Lewis @cghlewis.bsky.social · 1d

Thanks for checking the newsletter out! Ooof, I don't think I can choose a favorite because they all are very interesting and helpful for different reasons. But I think the AI-generated participant data article is one that probably piques a lot of interest right now.

Crystal Lewis @cghlewis.bsky.social · 1d

Issue 16 of RDM Weekly is out! 📬

It includes:
- Data is Not Available Upon Request @ianhussey.mmmdata.io
- AI Generated Participants in Social Science @jamiecummins.bsky.social @science.org
- Why’s it Hard to Teach Data Cleaning? @randyau.com
and more!

rdmweekly.substack.com/p/rdm-weekly...

RDM Weekly - Issue 016

A weekly roundup of Research Data Management resources.

rdmweekly.substack.com

2 7 18

Reposted by Crystal Lewis

Frank Hull @frankiethull.bsky.social · 1d

🥴

"300 TB of data

Before and after" meme

Shows breaking bad evolution

"Thinking about 300 TB vs looking at 300 TB meme"

Showing Samuel L Jackson in tux vs tank top and in-character with wild hair and crazy eyes

300TB of new data
300TB of new data
Meme

First face is happy second face is in despair lol

3 1 15

Reposted by Crystal Lewis

Kim Weeden @weedenkim.bsky.social · 1d

Just filled out a web survey with a bunch of Likert-type items. The response categories were in the same order on each item, but not the order you'd expect:

fairly important
important
unimportant
very important

Pretty sure "alphabetize response categories" is not best practice in survey design.

3 2 44

Crystal Lewis @cghlewis.bsky.social · 1d

When you've been working with someone for a while and you start to see the little ways that you are impacting how they work with data. 🤩

The name of a file someone just shared with me
"feedback_survey_raw_2025-08-15"

3 1 27

Reposted by Crystal Lewis

Berna Devezer @devezer.bsky.social · 1d

It's so deflating to lose an irreplaceable staff member. It's worse when you lose them to another unit on campus. I view that as a clear administrative failure and so should the admin. Academic staff is the glue that holds everything together yet they're so routinely underpaid and underappreciated.

1 4 37

Crystal Lewis @cghlewis.bsky.social · 1d

Oh no! 😅 I'm sorry, John!

Crystal Lewis @cghlewis.bsky.social · 1d

That is definitely a way to look on the bright side!

1 2

Crystal Lewis @cghlewis.bsky.social · 1d

Does it mean you're doing too much when you get the late start date wrong and you get your kiddo to school 1.5 hours late? 🤦

2 9

Reposted by Crystal Lewis

Dominique Baker @bakerdphd.bsky.social · 2d

"Deloitte Australia will issue a partial refund to the federal government after admitting that artificial intelligence had been used in the creation of a $440,000 report littered with errors including three nonexistent academic references and a made-up quote from a Federal Court judgement."

bianca wylie @biancawylie.com · 2d

“Deloitte was forced to investigate the report after University of Sydney academic Dr Christopher Rudge highlighted multiple errors in the document.”

www.afr.com/companies/pr...

Deloitte to refund government, admits using AI in $440k report

Deloitte will issue a partial refund to the government after admitting that artificial intelligence had been used in the creation of a report littered with errors.

www.afr.com

3 140 280

Crystal Lewis @cghlewis.bsky.social · 4d

Some Saturday reading 📖

cghlewis.com/blog/excel_e...

Tips for data entry in Excel | Crystal Lewis

This post provides a few tips for collecting higher quality and more usable data when using Excel as a data entry tool.

cghlewis.com

6 19

Crystal Lewis @cghlewis.bsky.social · 4d

You know you're watching something from the 90s when you hear the term "The Net".

3 18

Reposted by Crystal Lewis

Nick Tierney @njtierney.bsky.social · 5d

I'm giving a talk next week about my favourite thing: functions! Come along!

What: Practical Functions - Practically Magic
When: 8th/9th October - www.timeanddate.com/worldclock/f...
Where: Online, via Salt Lake City R User Group www.meetup.com/slc-rug/even...
How: @juliasilge.com

#rstats

Event Time Announcer - Salt Lake City talk: Practical Functions, Practically Magic

Event Time Announcer shows time for Salt Lake City talk: Practical Functions, Practically Magic in locations all over the world. In Salt Lake City it happens on Tuesday, October 7, 2025 at 4:00:00 pm.

www.timeanddate.com

1 4 17

Crystal Lewis @cghlewis.bsky.social · 5d

Prioritize documentation that has the biggest ROI for you, integrate documentation into your project workflow (assigning team members as responsible for it and setting aside times to update it), and also automate what you can (for instance versioning).

1 2

Crystal Lewis @cghlewis.bsky.social · 5d

I think teams know it takes time and they struggle to keep up with it. Also, some teams are just unsure how to get started with this type of documentation.

1 2

Crystal Lewis @cghlewis.bsky.social · 5d

The questions you ask are dependent on the data and the issues you run into. If you don't want to slow down a workflow, make sure you obtain all the documentation necessary to allow you to understand data lineage. Otherwise, be prepared to start asking those questions. :)

1 1 2

Crystal Lewis @cghlewis.bsky.social · 5d

If you want to be a good data manager, you have to get really comfortable with asking a lot of questions. When something is unclear or doesn't seem right, you can't settle or make assumptions. That's how you end up with bad data. Stay curious.

1 7 45