Skip to main content

Posts

Suicide hotline efficacy data: Assessment, descriptive data, t-tests, correlation, regression examples abound

ASIDE: THIS IS MY 500th POST. PLEASE CLAP. Efficacy data about a mental health intervention? Yes, please. The example has so much potential in a psych stats classroom. Or an abnormal/clinical classroom, or research methods. Maybe even human factors, because three numbers are easy to remember than 10? This post was inspired by an NPR story  by Rhitu Chatterjee. It is all about America's mental health emergency hotline's switch from a 10-digit phone number to the much easier-to-remember three digits (988), and the various ways that the government has measured the success of this change. How to use this (and related material) in class: 1) Assessment. In the NPR interview, the describe how several markers have improved: Wait times, dropped calls, etc.  Okay, so the NPR story sent me down a rabbit hole of looking for this data so we can use it in class. Here is the federal government's website about  988  and a link to their specific  988  performance data,...

Our World in Data's deep dive into human height. Examples abound.

Stats nerds: I'm warning your right now. This website is a rabbit hole for us, what with the interactive, customizable data visualizations. Please don't click on the links below if you need to grade or be with your kids or drive.  At a recent conference presentation, I was asked where non-Americans can find examples like the ones I share on my blog. I had a few ideas (data analytic firms located in other countries, data collected by the government), but wanted more from my answer.  BUT...I recently discovered this interactive from Our World in Data. It visualizes international data on human height, y'all  with so many different examples throughout. I know height data isn't the sexiest data, but your students can follow these examples, they can be used in a variety of different lessons, and you can download all of the data from the beautiful interactive charts. 1. Regressions can't predict forever. Trends plateau.  I'm using this graph to as an example of how a r...

Caffeine and Calories: An example of a non-linear relationship

Not all of our class examples should reject the null. Sometimes, you just need some non-significant data, small effect size data that doesn't detect a linear relationship. Such is the linear relationship between the number of calories and mg of caffeine in these 29 different treats provided by InformationIsBeautiful. InformationIsBeautiful provides that data , as do I .