A webring https://t.co/tBqNLdhnz0 - I added this blog to Randy Au's webring for data writers. Take a look here, or below: I'm struggling to describe why I think this is so interesting. Most of the writing I've seen about doing data work is misleading. It misrepresents what d…

The difference between a masters and Phd for data science is debugging import errors now vs. debugging import errors in 4 years.

SQL style - where do the commas go? https://t.co/PinI31HUho - TL;DR: there are good arguments for leading commas, but I recommend using trailing commas for consistency. We didn't take an opinionated stance on comma placement when writing Mozilla's SQL Style Guide probably to …

Talk: Practical Strategies for Data Storytelling https://t.co/bTmAjxbaVr - I gave a talk at the Open Data Science Conference a few weeks ago titled: Practical Strategies for Data Storytelling. I have more to say on the subject, but in the meantime you can see my slides and s…

@ejames_c Right, causal inference is what you do when you cannot apply an experimental treatment. But if the reason why you cannot is corporate org structure and internal politics, consider trying to address that head-on.

figured out how to make a reusable, linkable, animated component-based grammar of graphics with React + Framer Motion + D3 helpers the framer-specific stuff is now mostly an implementation detail within the components, o/w just useState & useContext

Some have noticed that I’m not tweeting much on the “Modern Data Stack” - I strongly believe that analytics/AE teams have a set of tools that are good enough for now (for BI, reporting etc). Let them settle; make an impact. Then let’s resume convo on what else you may need.

I’m going to be here at 3:30 ET talking data storytelling 🤘

The thing I find most frustrating about personal computing in 2022 is that no matter what you pick up your device to do, it'll try to convince you to do something else first. It's a constant battle of wills.

Our new study shows that data availability statements are not very useful; 1670 (93%) authors who indicated that data are available on request either did not respond or declined to share their data with us. Journal of Clinical Epidemiology: https://t.co/4IT2Dgphl4

Reminds me of a trick from "Influence": people are more likely to say yes if you give a reason, even if it's a bad reason. So, for example, you'll have better success asking "Can I cut in line because I'm in a rush" than just asking "Can I cut in line?"

This feels like a cognitive bias - like humans are more likely to believe something if it comes with a statistic, regardless of what the statistic says.

Nailed it. I hate when I see people adding a thin veneer of science to their argument. This ends up breeding an overall distrust in data.

"[the data] serves little purpose other than to provide an aura of scientific certainty that makes me, baselessly, more convinced that whatever narrative the story presents is true." - @bennstancil

A few years ago I had a manager who had never managed a Principal Eng before, and didn't understand how to measure impact outside of Jira tickets. So much of my work was invisible, preventing fires. I started sending him my brag doc every week! Thank you for the read, @harterrt

It's wild that any software works at all.

In vlookup veritas

I’m realizing slowly that I’m not meant to climb a career ladder. I’m meant to create my own.

My position today is that every engineer should maintain a brag doc. When your manager needs to ask for your promotion or get you a raise, you are helping them out by providing them supporting material and saving them SO MUCH TIME, and accelerating your own success.