Methods: Participants were recruited from a large predominantly White public university in the Midwestern United States. Using stratified random sampling, a total of 3,850 who self-identified as African American, Asian/Asian American, Latina/o American, or Native American were invited to participate as part of a larger racial microaggressions study. Among them, 1,710 students of color completed the survey during the 2011 – 2012 academic year. For the purposes of this study, we focused on the responses from all domestic undergraduate students who self-identified as Asian/Asian American (N = 244). The Artificial Neural Network (ANN) was conducted to verify variables that predict the level of depressive symptoms on the participants. The model with one hidden layer holding size hidden neurons showed the best performance.
Results: In order to create a training sample, 70.4% of the entire dataset was chosen through a random selection. A testing sample used 29.6%, and the remaining was used as a holdout sample. The error rate was approximately 17.8%, and the results from the sensitivity analysis, Receiver Operating Characteristics (ROC) curve, showed that the Area Under the Curve (AUC) was .807. The most powerful predicting variables in the neural network were experiencing racial microaggressions, students' ethnic identity, and the sense of belonging.
Conclusions and Implications: The study found that, consistently from the previous literature, experiencing racial microaggressions had a negative impact on individuals' mental health status. Moreover, the fact that participants' level of ethnic identity and sense of belonging had significantly high predicting ability was hopeful. Findings from the study emphasizes the importance of various efforts to support developing Asian students' sense of belonging in the campus community and cultivate their ethnic identity. The given study makes a meaningful contribution to the previous literature on Asian college students' experiences of racial microaggressions by applying a machine learning technique.