Skip to main content

Posts

Showing posts from May, 2026

MOAR GULL DATA!! Also, an actual independent t test and a conceptual factorial ANOVA.

TL;DR: Birds fly away from men a bit sooner than they fly away from women. Full stop. Here is the  original article,  and here is a write-up from  Nautilus . I love bird research. I'll get into why below. For now, let me show you how to use this example to teach three different lessons in a stats class. 1. Independent t test example with a data set The researchers shared their data. The researchers didn't analyze this data with a t test. But they did share this data visualization that looks a whole lot like one: Damn, I love the new trend of the box/violin/jitter plot. FYI: Researcher gender/the IV is labeled "gender," and how far the birds were before they flew away/the DV is labeled "FID" (flight initiation distance). Also, I love this example because the data violate the assumption of equal variance and provide a case for discussing Welch's test. 2. Conceptual example for Factorial ANOVA This example pairs well with a  previous blog post  featuring ...

Using GenAI to generate teachable data sets (here, an independent t test)

Two things I love to use when teaching stats are: 1)  Journal of the American Medical Association (JAMA) visual abstracts . I've blogged about them before. 2) Useful tools to generate pretend data sets that mimic real data, and use those pretend data sets to teach. See: Richard  Landers '  and Andrew Luttrell 's websites. So, I was delighted when I saw this recently posted visual abstract about  Ewing-Cobbs et al. (2026) research on using a specific CBT program to reduce stress in children following a traumatic physical injury .  https://jamanetwork.com/journals/jamapediatrics/fullarticle/2848163 I have a new  example of an independent  t  test  for class. Yay! And I teach tons of future nurses/PAs, so it is doubly applicable. However, the authors stated that the data wasn't immediately available. Also, once it is available, they (very reasonably) want to track their data sharing. Meaning that even if I could get their data, I shouldn't be s...