Data Scientist for Chesterfield Co Public Schools. Interested in education, data science, reading & writing, CrossFit, baking, and good dogs. ericekholm.comhttps://bsky.app/profile/ekholme.bsky.socialhttps://bsky.app/profile/ekholme.bsky.social/post/3mst2erxejk27Here’s the third (and probably final) post in a mini-series contextualizing AI applications within prominent ed psych research: AI tools are (mostly) great for retrieval practice https://www.edudata.blog/ai-tools-are-mostly-great-for-retrieval-practice/ #ai #education #edpsych11 Aug 2026 16:37 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3mst2erxejk27https://bsky.app/profile/ekholme.bsky.social/post/3moja6gfa7s2fNew blog post: AI probably inhibits flow states in students https://www.edudata.blog/ai-probably-inhibits-flow-states/ #education #edpsych #ai17 Jun 2026 21:13 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3moja6gfa7s2fhttps://bsky.app/profile/ekholme.bsky.social/post/3mgxoz5xqzs2hNew edudata blog post: embracing a Unix-like philosophy toward data www.edudata.blog/em/ #edusky #databs https://www.edudata.blog/em/13 Mar 2026 19:54 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3mgxoz5xqzs2hhttps://bsky.app/profile/ekholme.bsky.social/post/3me4wg5gjes2pNo matter how good I think my code is at the time of writing, I’m almost always repulsed by it 6 months later https://media.tenor.com/OUxqDVm0_78AAAAC/wolfhowling-yes.gif?hh=353&ww=49805 Feb 2026 17:35 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3me4wg5gjes2phttps://bsky.app/profile/ekholme.bsky.social/post/3m73tgkeoz224The more I code in languages other than #rstats, the less I like R’s preference for implicit returns. Sure, you can save yourself a line of code, but the explicit return just feels better03 Dec 2025 16:03 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3m73tgkeoz224https://bsky.app/profile/ekholme.bsky.social/post/3m4h3wh2wys2eNew edudata blog post: being “data driven” vs being “data informed” #edusky #databs https://www.edudata.blog/being-data-informed-vs-data-driven/30 Oct 2025 23:18 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3m4h3wh2wys2ehttps://bsky.app/profile/ekholme.bsky.social/post/3m3un3r6vzc25Most data science work is just asking SMEs to help me understand their data/help me make sure I’m filtering the data correctly.23 Oct 2025 15:05 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3m3un3r6vzc25https://bsky.app/profile/ekholme.bsky.social/post/3lzraapteo22mI just pushed v0.1.0 of my {blueycolors} #rstats package to GitHub. Check it out if you’re interested in using Bluey-themed colors in ggplot: ekholme.github.io/blueycolors/ https://ekholme.github.io/blueycolors/26 Sep 2025 19:47 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3lzraapteo22mhttps://bsky.app/profile/ekholme.bsky.social/post/3lwyqkbivz22tAfter a few week hiatus, edudata is back with another post. This one reviews Virginia’s fall-to-spring VALLSS growth report and offers some critiques about measuring student growth using ordinal categories: https://www.edudata.blog/virginias-vallss-2024-25-fall-spring-growth-report/?ref=edudata-newsletter #dataBS #edusky22 Aug 2025 15:11 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3lwyqkbivz22thttps://bsky.app/profile/ekholme.bsky.social/post/3lshp6k5sps24I just learned about “doubleML models.” Is this just data science/ML people trying to reinvent causal inference with the worst name possible? #rstats #datascience25 Jun 2025 22:51 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3lshp6k5sps24https://bsky.app/profile/ekholme.bsky.social/post/3ls7wmo2ows27This week’s Edudata blog post is on the narrative fallacy, school improvement, and root cause analysis https://www.edudata.blog/the-narrative-fallacy-and-education/ #education22 Jun 2025 20:43 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3ls7wmo2ows27https://bsky.app/profile/ekholme.bsky.social/post/3lqfeyt2xtk2xI’ve been doing stats/data stuff “seriously” for about 10 years, and I still think bootstrapping is voodoo. It’s the “ONE WEIRD TRICK” of statistics.30 May 2025 13:53 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3lqfeyt2xtk2xhttps://bsky.app/profile/ekholme.bsky.social/post/3lpd7dnzenk2wNew edudata blog post: alternatives to stacked bar charts, and some thoughts on tradeoffs when visualizing data #dataBS #rstats #edusky https://www.edudata.blog/stacked-bar-charts-and-data-viz-tradeoffs/16 May 2025 23:41 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3lpd7dnzenk2whttps://bsky.app/profile/ekholme.bsky.social/post/3lo7njsw4fc2vNew edudata blog post: In which I self-consciously offer a framework for data-driven decision making #datascience #edusky https://www.edudata.blog/some-hopefully-useful-questions-to-guide-data-driven-decisions/?ref=edudata-newsletter02 May 2025 20:19 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3lo7njsw4fc2vhttps://bsky.app/profile/ekholme.bsky.social/post/3lo22zx5ptc2yTFW the Google Gemini VSCode extension just writes your #rstats function documentation for you https://media.tenor.com/eA90OUd3y80AAAAC/dance-cheer.gif?hh=333&ww=49830 Apr 2025 15:05 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3lo22zx5ptc2yhttps://bsky.app/profile/ekholme.bsky.social/post/3lnnfqjivvs2uNew edudata blog post: what can educators learn from luxury brands like Hermes? #edusky #dataBS https://www.edudata.blog/luxury-goods-and-non-negotiable-values/25 Apr 2025 14:12 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3lnnfqjivvs2uhttps://bsky.app/profile/ekholme.bsky.social/post/3ln464wwbss2fNew edudata post: do we really need “big data” to tell us obvious things? https://www.edudata.blog/i-feel-like-im-taking-crazy-pills/?ref=edudata-newsletter18 Apr 2025 17:41 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3ln464wwbss2fhttps://bsky.app/profile/ekholme.bsky.social/post/3lmfndq4yrc2xI’m writing an #rstats package to work with data extracts (csvs) from a specific source. I mostly want a set of functions that will do data-frame-like operations on these extracts. My inclination is to create a new class that’s a thin wrapper around the data (formatted as a dataframe)…09 Apr 2025 18:42 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3lmfndq4yrc2xhttps://bsky.app/profile/ekholme.bsky.social/post/3llwd7sfy7c2mFor my educator friends in the RVA area, VCU is hosting (and I’m helping run) a free event on AI in education on 4/16.03 Apr 2025 16:30 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3llwd7sfy7c2mhttps://bsky.app/profile/ekholme.bsky.social/post/3llguxlyy5s2yThis week’s Edudata newsletter is a reaction to Emily Oster’s recent article on why we still need test scores: https://www.edudata.blog/reaction-to-emily-osters-why-we-need-test-scores/?ref=edudata-newsletter #dataBS #edusky28 Mar 2025 13:05 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3llguxlyy5s2yhttps://bsky.app/profile/ekholme.bsky.social/post/3lkvte7muns2cThis week’s Edudata newsletter is about writing, AI, and owning your thinking: https://www.edudata.blog/writing-for-busy-readers-and-why-not-to-use-ai/?ref=edudata-newsletter #dataBS #edusky21 Mar 2025 18:21 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3lkvte7muns2chttps://bsky.app/profile/ekholme.bsky.social/post/3lkeezw3vy227New edudata post: when do predictions about students’ future performance help? https://www.edudata.blog/when-do-predictions-about-students-future-performance-help/?ref=edudata-newsletter #edusky #databs14 Mar 2025 19:50 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3lkeezw3vy227https://bsky.app/profile/ekholme.bsky.social/post/3ljsi5k57d223New edudata blog post: does your data structure match the question you want to answer? https://www.edudata.blog/does-your-datas-structure-match-the-question-you-want-to-answer/?ref=edudata-newsletter #edusky #databs07 Mar 2025 16:58 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3ljsi5k57d223https://bsky.app/profile/ekholme.bsky.social/post/3ljqeja3o4c25I haven’t done #TidyTuesday in years apparently, but I decided to fiddle with this week’s dataset to play with the S7 system in #rstats https://github.com/ekholme/TidyTuesday/blob/master/2025_9_animal_shelter/script.R I haven’t used it enough to have an opinion, but I’m excited to see where this goes!06 Mar 2025 20:47 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3ljqeja3o4c25https://bsky.app/profile/ekholme.bsky.social/post/3ljq6fu4wcc2eAnyone have resources/suggestions on including spatial data (lat/long) in a propensity score model? My gut is to use a non-parametric model (eg random forest) to predict intervention exposure and then use those predictions to get weights. But spatial data is not my expertise #databs #rstats06 Mar 2025 18:58 +0000at://did:plc:vvy5k34fpy2afmu3ulx2guw7/app.bsky.feed.post/3ljq6fu4wcc2e