Measuring change is our first ever LIVE EPISODE! Recorded at the Association for Psychological Science conference with a live conference audience! Over here in the developmental sciences, we are often trying to measure or predict how much people grow and change over time. To do that, we’ll sometimes measure a skill or ability twice (for kids, maybe that’s once near the beginning of the school year and once at the end of the school year). In this episode, Jess and Sara talk about options for statistically modeling and predicting change between two timepoints. We talk about two basic models: the simple difference scores and residualized gain scores, explain why we can’t just measure people at post test, and conclude that the question is important in making your decision. Have a listen as we get to take some related questions from the audience.
Things we mentioned:
If you are working in the latent space and have a measurement model at each time point, you may need to establish measurement invariance across time. To do so, you can follow steps in Brown (2015; pp. 259–265).
Brown, T. A. (2015). Confirmatory factor analysis for applied research (2nd ed.). Guilford Publications.
Connect with the podcast on twitter @within_between, or email us letters about developmental science at withinandbetweenpod@gmail.com. More episodes and podcast information at WithinandBetweenPod.com. Follow Dr. Hart on twitter @Saraannhart Follow Dr. Logan on twitter @Jarlogan. Our theme music was composed by Jason Flowers. Our logo was created by Nathan Archer. Recorded May 27th, 2023.
