Although text data are ubiquitous in organizations, the advancement of text analysis methods has created an unfortunate bottleneck for many organizational researchers. This paper introduced the Structure Text Model (STM; Roberts et al., 2014), a cutting-edge text modeling method that uses an unsupervised machine learning model to statistically derive latent semantic topics underlying a collection of text documents. More importantly, the STM method has the advantage of modeling document-level variables (i.e., metadata) as covariates, which is critical for organizational research. We also demonstrated the application of STM in diversity research: comprehensively analyzing a large number of diversity statements publicly released by Fortune 1000 companies. Our structural topic modeling uncovered six underlying latent semantic topics: 1) general DEI terms, 2) supporting Black community, 3) acknowledging Black community, 4) committing to diversifying workforce, 5) miscellaneous words, and 6) titles and companies. We further explored and found that the prevalence of these topics varied as a function of company characteristics, including industry sector, CEO race, corporate political orientation, etc. Our paper not only demonstrates the promising application of Structural Text Models in organizational research, but also provides important theoretical implications for current diversity research through the meaningful findings.