TL;DR: This cool, interactive website asks you to participate in a replication. It also explains how a researcher decision on how to define "randomness" may have driven the main effect of the whole study. There is also a scatter plot and a regression line, talk of probability, and replication of a cognitive example. Long Version:  This example is equal parts stats and RM. I imagine that it can be used in several different ways: -Introduce the replication crisis by participating in a wee replication -Introduce a respectful replication based on the interpretation of the outcome variable  -Data visualization and scatterplots -Probability -Aging research Okay, so this interactive story from The Pudding is a deep dive into how one researcher's decision may be responsible for the study's main effect.  Gauvrit et al. (2017 ) argue that younger people generate more random responses to several probability tasks. From this, the authors conclude that human behavioral complexity...

Sometimes, I post whole class lessons. Sometimes, I post short little example nuggets. Today I share the latter.  This one is a brief, easy-to-understand example of why AI only learns what we teach it and how even a smarty pants computer can get a little confused about correlations and what they mean. A great way to introduce ML, AI, problems with both, and even discuss correlation and predictions and regression. https://twitter.com/hoalycu/status/1507770891786096643...in my head, I imagine that AI was just judging comic-sans font. The text in this tweet was from a MIT Technology Review article by WIll Douglas Heaven: https://www.technologyreview.com/2021/07/30/1030329/machine-learning-ai-failed-covid-hospital-diagnosis-pandemic/ If you want to go deeper with this example, I strongly recommend reading Dr. Cat Hicks's thread about this post:  https://twitter.com/grimalkina/status/1508095358693302275 .

A bit over a year ago, Susan Nolan asked me to chair the Statistical Literacy, Reasoning, and Thinking, Guidelines 2.0 for the Society for the Teaching of Psychology.  We were asked to explore and provide guidance for a) teaching statistical thinking in intro psychology and b) understanding how statistical thinking is taught across the psychology curriculum. This post will highlight the accomplishments of the first group, which created easy-to-implement teaching exercises that emphasize statistical reasoning skills in Intro Psych. The Guidelines 1.0 group provided lists of topics included in Intro Psych. The Guidelines 2.0 convened and created a series of brief, easy-to-apply exercises that correspond to the core topics typically taught in Intro. The sub-committee chair, Dr. Garth Neufeld, shared his considerable expertise about Intro Psychology to lead the group and center each exercise in American Psychological Association and American Statistical Association guidelines for under...

 Sophie Hill created a great game that shows students how to "eyeball" regression lines (or just lines) by guessing the y-intercept and the slope.  At the beginning of the game, you get a scatter plot. Then, you need to guess the y-intercept and the slope.   Once you make a guess, it will show you the actual line of best fit...and your line, along with residuals and mean squared error. So, this doesn't just allow for eyeballing the regression line but also how to test the fit of a line. P.S.: If you liked this, you'd love the Guess the Correlation game.

 My favorite examples of bad data visualizations are the ones that use accurate data that was actually collected through seemingly ethical means but totally malign the data. The numbers are correct, the data viz is...not very truthy ( I'm looking at you, Florida. ) Especially when you mess up the data viz in a way that appears to be deliberate AND doesn't really strengthen your point. I'm also looking at you, The White House. Here is a story of a deliberate but pointless massaging of a y-axis. A story in Three Tweets. 1. The Biden Administration is doing a good job of encouraging economic growth, right? Take a gander at this bar graph. 2021 was a success...just look at the chart.  2. BUT WAIT. What's this? That y-axis is shady. I...just can't think of any software/glitch that could make this mistake by accident. ALSO: If you like Twitter, follow Graph Crimes.  3. The White House issues a correction featuring a pretty good data put, I would say.  FIN

Alright, this is a 30-second long example for a) bimodal distributions and b) why measures of variability matter when we are trying to understand a mean. And that mean is...AGE OF DEATH. My inspiration for this tweet is: I’m just a girl, standing in front of the internet, asking it to understand that historical life expectancies doesn’t mean most people died at 45 but rather that infant mortality was super high and pulled down the average. — Angelle Haney Gullett (@CityofAngelle) January 12, 2022 Gullett refers here to the commonly held belief that if the mean life span Back In The Day was 45, or thereabout, everyone was dying around 45. NOT SO. Why? The short answer is no. Broadly speaking, there were two choke point of human mortality. Younger than 5, and again around 50. If you made it through those, barring accidents, you likely had what was a normal lifespan of ~65-70 years. And this is why I’m no fun at parties 😂 — Angelle Haney Gullett (@CityofAngelle) January 12, 2022 OK. An...

Many thanks to my colleague, Andrew Caswell, for sharing this Reddit post with me: https://www.reddit.com/r/dataisbeautiful/comments/s75sm7/oc_us_life_expectancy_vs_of_taco_bell_locations/ So, this alone is an excellent example of correlation and the third variable problem. But...more delightfully, the Redditor who created this graph also shared where he found this data (https://www.nicerx.com/fast-food-capitals/, https://worldpopulationreview.com/state-rankings/life-expectancy-by-state). BETTER STILL: I downloaded and organized all of the fast-food data and mortality data and put it in one spreadsheet for you all. Do All The Correlations! Teach your students about Bonferroni corrections! Figure out the fast-food restaurant that correlates the most strongly with mortality!   PS: Did you know that there is an option to download data from a website in Excel?  The fast-food data was presented in an embedded, scrolly table, and that Excel option made it easy-peasy to do...