iiinova — where your digital self keeps evolving

Reposted by Crystal Lewis

Crystal Lewis @cghlewis.bsky.social · 4h

Data Cleaning for Data Sharing 3 Part Tutorial (from Love Data Week 2023). Part 1: cghlewis.com/blog/data_cl... Part 2: cghlewis.com/blog/data_cl... Part 3: cghlewis.com/blog/data_cl...

041

Crystal Lewis @cghlewis.bsky.social · 2h

Replacing my laptop battery myself for the 1st time and hoping I don’t kill my computer. Pray for me.

0100

Crystal Lewis @cghlewis.bsky.social · 4h

Data Cleaning for Data Sharing 3 Part Tutorial (from Love Data Week 2023). Part 1: cghlewis.com/blog/data_cl... Part 2: cghlewis.com/blog/data_cl... Part 3: cghlewis.com/blog/data_cl...

041

Crystal Lewis @cghlewis.bsky.social · 13/10/2026

I’m tired of everything I come across being about AI or made by AI. 😫

1200

Crystal Lewis @cghlewis.bsky.social · 13/10/2026

Issue 66 of #rdmweekly is out! 📬 ➡️ What We Learned Trying to Reproduce Psychological Science @i4replication.bsky.social ➡️ Data Everywhere, But are They Fit for Purpose? @science.org ➡️ An Introduction to OpenRefine @mcgill.ca and more! rdmweekly.substack.com/p/rdm-weekly...

rdmweekly.substack.com

RDM Weekly - Issue 066

A weekly roundup of Research Data Management resources.

0145

Crystal Lewis @cghlewis.bsky.social · 09/10/2026

Hi! Yes, I've come across it before actually, and have shared it! Thanks for bringing it back to my attention. Great product.

110

Crystal Lewis @cghlewis.bsky.social · 08/10/2026

If your files look like this, there is help! 👇 datamgmtinedresearch.com/style

0113

Crystal Lewis @cghlewis.bsky.social · 07/10/2026

There is help out there. 😅

100

Crystal Lewis @cghlewis.bsky.social · 07/10/2026

100% But the difficult thing about choosing -99 in REDCap is that when you have a select all variable, you end up with variables that look like this race_1, race_2, race_-99 Not great. So maybe in those cases, 999 might be better

120

Crystal Lewis @cghlewis.bsky.social · 06/10/2026

Did you know that if you use REDCap project-wide missing value codes (i.e., 77 = I don't know), and that gets applied to a variable like age and someone's age is 77, when you export that data as a CSV (labels) REDCap converts those values to blank? Really gotta choose those missing values wisely!

2200

Reposted by Crystal Lewis

Hadley Wickham @hadley.nz · 06/10/2026

I’m not claiming that this video is particularly good, but this was ~1 minute of prompting, ~$2 of google apis, ~$3 of opus 5.5, and 10 minutes of waiting. I think if you put in even 15 minutes of prompting you could create something solid.

5537

Crystal Lewis @cghlewis.bsky.social · 06/10/2026

Issue 65 of #rdmweekly is out! 📬 It includes: ➡️ Life After OSF: How and Where to Share Data and Materials @syeducation.bsky.social ➡️ Tidy Data @datafrosch.bsky.social ➡️ What Even Are Parquet Files? @niccrane.bsky.social and more! rdmweekly.substack.com/p/rdm-weekly...

rdmweekly.substack.com

RDM Weekly - Issue 065

A weekly roundup of Research Data Management resources.

1146

Crystal Lewis @cghlewis.bsky.social · 05/10/2026

I wish I had come up with this title for a conference/conference presentation. Love it. 👻

0110

Reposted by Crystal Lewis

Jon Harmon (he/him/his) @jonthegeek.com · 05/10/2026

TIL (via a #TidyTuesday submission): In `dplyr::rename()`, the RHS of each = can be a column *number* (index). That makes sense, but it's documented as "new_name = old_name syntax", so I hadn't thought about the implications of what "old_name" can be!

4102

Crystal Lewis @cghlewis.bsky.social · 05/10/2026

ICYMI! RDM (Research Data Management) Weekly hit 1,000 subscribers over the weekend! Thanks to everyone who helped the newsletter reach this milestone by supporting and spreading the word about this free data management resource! #rdmweekly rdmweekly.substack.com

0173

Reposted by Crystal Lewis

Nicola Rennie @nrennie.bsky.social · 01/10/2026

If you missed today's @rinpharma.bsky.social workshop, here are a few of my favourite "quick fixes" for some common #ggplot2 issues 📊 1. Use {ggview} to preview plots at the same size and resolution as you want to save them with ggsave() to avoid the annoying resizing #RStats #DataViz

library(ggview)
ggplot(mtcars, aes(wt, mpg)) +
geom_point() +
canvas(
width = 6, height = 4,
units = "in",
dpi = 300
)

113143

Crystal Lewis @cghlewis.bsky.social · 04/10/2026

Some exciting Sunday news. RDM Weekly hit 1k subscribers! 🎉 That's 1,000 people accessing free data management resources every week! rdmweekly.substack.com #rdmweekly

0120

Reposted by Crystal Lewis

samuel mehr @mehr.nz · 02/10/2026

early-career academics, if you don't have a website, make yourself one. yes it's annoying but just do it it can be plain text on a white background, or google sites, or even a free Wix one with ads. nobody cares we just need to be able to look you up, quickly, bc there's like 100 other apps here

1119058

Crystal Lewis @cghlewis.bsky.social · 01/10/2026

I'm not crying, you're crying. 🥹 So grateful to have received this from the POWER (Providing Opportunities for Women in Education Research) leadership team.

7892

Crystal Lewis @cghlewis.bsky.social · 30/09/2026

Same. I spent 3 years writing my book (on my own time). The financial pay off is small, partly because my book is OA, but I'm still glad I did it. It let me synthesize and document my knowledge and experiences and I feel like it's making a difference. Takeaway: Write for you, not for the money.

091

Crystal Lewis @cghlewis.bsky.social · 29/09/2026

Issue 64 of #rdmweekly is out! 📬 It includes: ➡️ Applied Data Skills @psyteachr.bsky.social ➡️ Excel Now Supports Multiple Values in a Single Cell ➡️ NIH Details $170 Million Biomedical Research Reproducibility Effort @science.org and more! rdmweekly.substack.com/p/rdm-weekly...

rdmweekly.substack.com

RDM Weekly - Issue 064

A weekly roundup of Research Data Management resources.

075

Crystal Lewis @cghlewis.bsky.social · 28/09/2026

Changing measure items midway during a project.

0122

Crystal Lewis @cghlewis.bsky.social · 27/09/2026

23 subscribers away from 1k! Help #rdmweekly meet that milestone! 🎉 rdmweekly.substack.com

0161

Crystal Lewis @cghlewis.bsky.social · 25/09/2026

I haven't tried it yet. I don't think the Excel functionality is even available to most people yet. And I believe JSON has its own import package.

000

Crystal Lewis @cghlewis.bsky.social · 25/09/2026

I saw this yesterday, and I thought, yes it violates tidy data. But also I see the benefit for some people who are working in Excel only. And you can do similar things in R (nested list columns). But yeah, I'm sure it will ultimately just create more headaches in the research world. :/

3124

Crystal Lewis @cghlewis.bsky.social · 25/09/2026

I think that is a common theme at tech conferences right now.

100

Crystal Lewis @cghlewis.bsky.social · 25/09/2026

AI: Getting more and more advanced, things progressing quickly Me helping researchers: Let's talk about the benefits of documenting any work that you do, or naming things so that you understand them later.

static.klipy.com

The Simpsons: Marge Simpson Go-Karting

ALT: The Simpsons: Marge Simpson Go-Karting

3376

Crystal Lewis @cghlewis.bsky.social · 25/09/2026

Also, I was just checking out your resources and I'm excited to share them in my newsletter as well!

110

Crystal Lewis @cghlewis.bsky.social · 25/09/2026

I have several resources! - datamgmtinedresearch.com/clean - cghlewis.com/talks/ncme/ - github.com/Cghlewis/dat... - osf.io/e5g6t/files/... But also if you search my newsletter for "data cleaning" I think I've shared several from others there! rdmweekly.substack.com

110

Reposted by Crystal Lewis

Heather Urry @heatherurry.io · 24/09/2026

holy fucking shit - see Claude Code (Opus 5.5) rip through a survey eligility test I was really proud of... It requires tracing a route through designated points on a map with server-side Javascript applying eligibility criteria. (1.5 min video)

63522

Crystal Lewis @cghlewis.bsky.social · 24/09/2026

Totally fair! I'm surrounded by PhDs so it's a good reminder!

010

Crystal Lewis @cghlewis.bsky.social · 23/09/2026

It's still in the works, but it's looking like I'll be giving my second keynote address, for one of my favorite events....Love Data Week. More to come.

static.klipy.com

Homer Simpson Says Woo Hoo Excited

ALT: Homer Simpson Says Woo Hoo Excited

4432

Crystal Lewis @cghlewis.bsky.social · 22/09/2026

Issue 63 of #rdmweekly is out! 📬 ➡️ @datadryad.bsky.social Gives Researchers due CRediT ➡️ Shared Research Materials in Psychology are Widely Downloaded but Often Difficult to Reuse @dsquintana.bsky.social ➡️ Intro to Vibe Coding in R & RStudio and more! rdmweekly.substack.com/p/rdm-weekly...

rdmweekly.substack.com

RDM Weekly - Issue 063

A weekly roundup of Research Data Management resources.

0133

Crystal Lewis @cghlewis.bsky.social · 21/09/2026

Either the LLM sees real data to build the synthetic version (which still exposes PII) or I have to characterize the real data's structure well enough to synthesize it. And most of the education datasets I work with have such messy edge cases, this would just not save me time.

120

Crystal Lewis @cghlewis.bsky.social · 21/09/2026

Almost all of my work involves cleaning data that contains PII. So I don't use LLMs for that.

120

Crystal Lewis @cghlewis.bsky.social · 18/09/2026

I was recently given a raw data file. I needed to recreate a cleaning process that was originally done manually several years ago. That team had written a very thorough README about that cleaning process. While it was tedious, I was able to recreate that process. Documentation pays off people.

1411

Crystal Lewis @cghlewis.bsky.social · 17/09/2026

Hey you. The one working with data right now. This is your periodic reminder to check to make sure you don't have duplicates in your data. Did someone submit 2 surveys when they should have only submitted one? Did someone's data get duplicated during a merge? Shit happens. Check your data.

1584

Crystal Lewis @cghlewis.bsky.social · 17/09/2026

Oooof, as if I wasn't already bummed enough that I couldn't make it this year. Now I'm really feeling it. 😅

150

Reposted by Crystal Lewis

Nicola Rennie @nrennie.bsky.social · 17/09/2026

I love overly honest READMEs

Screenshot of GitHub README saying known issues: coding style only a mother could love

19920

Crystal Lewis @cghlewis.bsky.social · 17/09/2026

Benefits of cleaning every wave: 1. Still fresh in your mind. 2. Allows you to catch any errors in your collection method that you can correct for future waves 3. May provide the opportunity to recollect bad data 4. The data is ready when you need it! datamgmtinedresearch.com/clean

datamgmtinedresearch.com

Chapter 14 Data Cleaning | Data Management in Large-Scale Education Research

Figure 14.1: Data cleaning in the research project life cycle. Even with the most well-designed data collection and capture efforts, data still require at least some additional processing before...

060

Crystal Lewis @cghlewis.bsky.social · 17/09/2026

Collecting longitudinal data? Know when you should start cleaning it? ❌ Not at the end of your project. ✔️ After every wave of collection See reasons why below. 👇

1213