Psychological and sociocultural patterns of cosmetic-procedure consideration: A machine-learning approach.

Cosmetic-procedure consideration is associated with a spectrum of psychological and sociocultural factors. However, traditional linear models are typically insufficient to capture the complex, non-linear, and interactive patterns inherent in these associations. To bridge this gap, the present study utilized machine-learning (ML) approaches-specifically decision tree and random forest models-to map the multivariable dynamics of cosmetic-procedure consideration. The sample comprised 912 Iranian adults recruited via online convenience sampling (60.3% female; Mage = 34.3 years), categorized into procedure-experienced (n = 281) and procedure-naïve (n = 631) groups. Decision-tree analyses showed distinct hierarchical pathways. While body dysmorphic concerns were the primary predictor for the procedure-naïve group, body shame served as the initial and most influential driver for those with prior cosmetic procedure experience. Random-forest models corroborated these findings, offering robust estimates of variable importance. Across both groups, self-objectification, social appearance anxiety, and body image evaluations emerged as critical predictors. These findings underscore the necessity of employing non-linear modeling to uncover subgroup-specific pathways, providing a more nuanced understanding of the psychological architecture underlying cosmetic-procedure consideration.