Athletic identity and depressive symptom risk in college students: a machine learning approach to identify at-risk profiles and protective correlates | Deqing Chang | ResearchPod

Transcript: Athletic identity and depressive symptom risk in college students: a machine learning approach to identify at-risk profiles and protective correlates

[[RP_SECTION:athletic-identity-and-mental-health|Athletic Identity and Mental Health]]

Alex: [measured, clear] The headline finding here is that athletic participation is a poor predictor of mental health. What actually matters is the identity an athlete builds around their sport—specifically, whether they're oriented toward mastery or toward status.

Sam: [curious] So the sport itself isn't the variable? That seems counterintuitive if we assume team sports provide protective social support.

Alex: [steady] Right. The data suggests that if you define yourself through status, popularity, and dominance—what the authors call a "jock identity"—you carry a significantly elevated risk of depressive symptoms, and training volume doesn't change that picture.

Sam: How did they actually separate those groups? [[RP_SECTION:methodological-approach-and-clustering|Methodological Approach and Clustering]]

Alex: [slower, for clarity] They ran K-means clustering on five identity constructs, including task and ego orientation. That move is important—it lets them step past simple participation counts and identify four distinct psychological phenotypes in the data. You can't recover those profiles from a standard regression because it would just average across them.

Sam: And then how do they connect the phenotypes to depression outcomes? [[RP_SECTION:shap-values-and-model-discovery|SHAP Values and Model Discovery]]

Alex: They trained gradient boosting classifiers and used SHAP values to decompose feature importance. The "jock" phenotype came out as a primary driver of elevated depression risk. The reason to use SHAP here isn't raw predictive accuracy—the authors are transparent that the AUC is nearly identical to a well-tuned logistic regression. The value is in discovery: the model surfaces interaction terms and non-linear relationships you'd otherwise miss, which you can then formalize into something more interpretable.

Sam: So it's functioning more as a search tool than a prediction engine.

Alex: Exactly. It gives you a template—here's where the interesting structure lives in this data—that a variable-centered model would have smoothed over.

Sam: Was the depression risk uniform across the jock phenotype, or were there moderators? [[RP_SECTION:gender-and-identity-moderators|Gender and Identity Moderators]]

Alex: Sex was a significant one. The association between status-seeking and depression was notably stronger among male students. The likely mechanism is that status loss in sport is perceived as a more direct threat to masculine legitimacy—it connects to the precarious manhood literature. If your self-worth is load-bearing on your social rank, any competitive failure becomes an identity-level threat, not just a bad game.

Athletic identity and depressive symptom risk in college students: a machine learning approach to identify at-risk profiles and protective correlates

Deqing Chang

Jun 5, 2026·4:14·zotero·Original Paper

Athletic IdentityMachine Learning in PsychologyJock vs. Athlete IdentityDepressive Symptom RiskJock Identity vs. Athlete IdentityPrecarious Manhood

Listen

4 min

Briefing

Research Overview

This study challenges the common assumption that sports participation is universally beneficial for mental health. By analyzing survey data from 795 college students, the researcher investigated how different ways of constructing an athletic identity—ranging from status-seeking to skill-mastery—relate to depressive symptom risk. Using machine learning, the study identified distinct psychological profiles and examined how factors like gender norms and risky behaviors influence these associations.

Athletic Identity Profiles

Using clustering algorithms, the study identified four distinct profiles of athletic identity:

Non-Athletes: Uniformly low scores across all athletic identity measures.
Recreational Athletes: Characterized by high task orientation and low jock identity.
Competitive Athletes: High athlete identity with balanced task and ego orientations.
Status-Oriented Athletes: High jock identity and ego orientation, with low task orientation.

These profiles showed significant differences in mental health outcomes. Status-Oriented Athletes exhibited the highest prevalence of elevated depressive symptoms (45%), while Recreational Athletes showed the lowest (24%).

Machine Learning and Risk Prediction

To predict depression risk, the study employed gradient boosting models (XGBoost). The models achieved an AUC of 0.78, demonstrating that athletic identity and motivational orientations provide meaningful predictive power beyond standard demographic data. Feature importance analysis revealed that conformity to masculine norms and risky drinking behaviors were the strongest predictors of depressive symptoms. Furthermore, the association between jock identity and depression was significantly stronger among male students, suggesting that sports serve as a high-stakes arena for masculine identity validation.

Why It Matters

These findings suggest that collegiate athletic programs should move away from a one-size-fits-all approach to student wellbeing. By identifying students who prioritize status over mastery, universities can implement targeted interventions—such as fostering mastery-oriented coaching climates and addressing the underlying pressures of restrictive masculine norms—to better support student mental health.

Key Terms and Definitions

Athletic Identity: The degree to which an individual defines themselves through their role as an athlete.
Jock Identity: A form of athletic identity centered on social status, popularity, and dominance rather than skill development.
Task Orientation: A motivational disposition focused on personal mastery, skill improvement, and intrinsic enjoyment of the sport.
Ego Orientation: A motivational disposition focused on demonstrating superiority over others and achieving social status.
SHAP Analysis: A machine learning interpretation technique used to quantify the contribution of each feature to the model's predictions and identify complex interactions between variables.

Abstract

The relationship between athletic participation and depressive symptom risk among college students remains complex and contested, with prior research yieldin...

Athletic identity and depressive symptom risk in college students: a machine learning approach to identify at-risk profiles and protective correlates: Research Paper Podcast