KMAN Publication Inc. (KMANPUB) Psychology of Woman Journal 3041-8615 7 Serial Number 28 2026 10 01 Predicting Body Dissatisfaction Among Young Women Using Random Forest Based on Social Media, Appearance, and Self-Evaluation Features Predicting Body Dissatisfaction Among Young Women Using Random Forest Based on Social Media, Appearance, and Self-Evaluation Features 1 17 EN Journal Article 2026 08 12 <table> <tbody> <tr> <td> <p><strong>Objective:</strong> This study aimed to predict body dissatisfaction among young women in Armenia using a Random Forest model based on social media-related, appearance-related, self-evaluation, demographic, and body-related features.</p> <p><strong>Methods and Materials:</strong> This quantitative cross-sectional predictive study included 412 young women aged 18–29 years residing in Armenia. Participants were recruited through universities, workplaces, community settings, and online platforms. Body dissatisfaction was assessed using the Body Shape Questionnaire-16A, while predictor variables included daily social media use, appearance-focused content exposure, social media appearance comparison, photograph editing and filtering, importance of online feedback, general appearance comparison, perceived attractiveness, body-shape satisfaction, current–ideal appearance discrepancy, appearance-related social pressure, global self-esteem, appearance-contingent self-evaluation, and body mass index. A Random Forest regression model was developed using an 80% training and 20% independent test split. Five-fold cross-validation was used for hyperparameter optimization. Model performance was evaluated using R², mean absolute error (MAE), and root mean squared error (RMSE), and permutation feature importance was used to identify the strongest predictors.</p> <p><strong>Findings:</strong> The Random Forest model demonstrated satisfactory predictive accuracy, explaining 61% of the variance in body dissatisfaction during cross-validation and 64% in the independent test sample (R² = .64, MAE = 7.21, RMSE = 9.36). General appearance comparison was the strongest predictor, producing a .119 decrease in R² after permutation, followed by body-shape satisfaction (.104), appearance-contingent self-evaluation (.092), social media appearance comparison (.081), current–ideal appearance discrepancy (.069), and global self-esteem (.057). Partial dependence analyses indicated nonlinear increases in predicted body dissatisfaction at moderate-to-high levels of appearance comparison, appearance-contingent self-evaluation, and social media appearance comparison, while greater body-shape satisfaction was associated with lower predicted dissatisfaction.</p> <p><strong>Conclusion:</strong> Body dissatisfaction in women can be predicted with substantial accuracy from an integrated set of appearance, self-evaluation, and social media-related characteristics, with psychological comparison and subjective appearance evaluation contributing more strongly than general social media exposure or body size alone.</p> </td> </tr> </tbody> </table> https://journals.kmanpub.com/index.php/psywoman/article/download/6196/11892