Exploring group drumming’s effects on stress biomarkers and social well-being among college students

Methods

Summary

This project investigates the immediate effects of group drumming on psychological and physiological stress indicators among college students. Participants, all students at Lynn University aged 18 or older, will take part in a 45-minute Adlerian-based group drumming session: a socially focused drumming activity based on Adlerian principles of connection and cooperation. Before and after the session, we will collect survey data on mood, social interest, and physiological data, including heart rate, blood pressure, and saliva samples. We will analyze saliva samples using ELISA assays for a panel of stress-related biomarkers, including cortisol, salivary alpha-amylase, immunoglobulin A (IgA), C-reactive protein (CRP), dehydroepiandrosterone (DHEA), oxytocin, and glucose. By integrating self-report measures with objective physiological biomarkers and applying principal component analysis (PCA) and pairwise comparison (before vs. after), our study aims to identify patterns of individual variation in response to the drumming intervention. The ultimate goal is to understand how musical group engagement may buffer stress, promote well-being in college students, and identify which students benefit most from this rhythmic, community-based approach.

Challenges

A key challenge in this study is managing individual variability in physiological and emotional responses to the drumming intervention. While biomarkers like cortisol, DHEA, and CRP provide objective insight into stress-related changes, they are sensitive to numerous confounding factors such as recent food intake, caffeine, nicotine use, and oral health. To address this, participants will receive detailed, visually guided pre-session instructions and reminders, and noncompliant individuals will be flagged during data collection for sensitivity analyses. To minimize inter-rater variability, all research assistants will undergo structured training and use scripted protocols for sample collection and survey administration. Given early findings that age, academic status, and prior music experience influence response patterns, these variables will be recorded and included as covariates or grouping factors in statistical models. Lastly, we will use principal component analysis (PCA) to explore multidimensional patterns to reconcile discrepancies between physiological and self-report outcomes, and exploratory subgroup analyses will help elucidate divergent response trajectories. These steps aim to ensure data quality, interpretability, and replicability.

Pre Analysis Plan

Initial analyses will employ paired-sample t-tests to assess pre/post differences in mood, social interest, and individual biomarkers (e.g., cortisol, alpha-amylase, oxytocin). To better understand multidimensional patterns of change, we will apply Principal Component Analysis (PCA) across change scores from psychological surveys and physiological markers. PCA will allow us to uncover latent structures representing distinct participant response profiles. We will then assess how these components vary as a function of demographic moderators such as age, academic classification, and prior music experience using non-parametric tests (Kruskal-Wallis and Mann-Whitney U) and effect size estimates (Cohen’s d). With this data analysis workflow, we seek to identify which student subgroups respond most strongly to group drumming and through which mechanisms (emotional, physiological, or both).