Happy Halloween, and welcome to the first new blog post and episode of Music Therapy Journal Club with Janice Lindstrom and Meganne Masko!
As always, our goal is to make research a little less scary and more applicable to clinicians and students. We hope you find this a treat rather than a trick (I promise that’s the last of the Halloween references).
You can listen to the podcast here: Music Therapy Journal Club 10/31/17
Today, we talked about the Summer 2017 edition of the Journal of Music Therapy (JMT, Volume 54, Number 2) published by the American Music Therapy Association. This edition of JMT had several research-specific terms, some of which may be unfamiliar to readers, so here is my vocabulary reference list.
Parallel Cluster-Randomized Controlled Trial (p. 136): The graphic on p. 142 provides a visual representation of the cluster randomized trial. In a PCRCT, is the cluster of people (rather than individuals) who are randomized to either the experimental or control condition. For more information on this design, see Hemming et al. (2017).
Effect Size (p. 139): There are statistics that tell readers if there are differences in test scores or other outcomes between two groups in a research study; however, those statistics don’t tell us how significant those differences are. That’s where the effect size comes into play. For more information about effect sizes and why they are important, see Sullivan & Feinn (2012).
LOCF Analysis (p. 150): According to Xu (2009), Last Observation Carried Forward is a computational method used to account for drop outs in a research using repeated measures. Rather than completely removing the drop out’s data from the statistical analysis (which can cause different statistical challenges), statisticians use the last observed value in all computations of data occurring after the person dropped out of the study. For example (see below), if Participant A was supposed to complete four different measures in a study, but they only completed the first two, we would keep using the value from data collection two to stand in for data points three and four rather than taking them out of the data pool.
Questions or comments? Please feel free to leave us a note, or email me (Meganne Masko) at email@example.com.