Knowledge-sharing Behavior on Facebook Using Theory of Planned Behavior

Knowledge-sharing behavior on Facebook gained new relevance during the COVID-19 pandemic, when face-to-face interaction was restricted and much of this activity moved online. This study analyzes such behavior using the Theory of Planned Behavior (TPB), pairing survey-based attributes with observed Facebook activity from 1,841 lengthy posts by 81 users. Twenty-one TPB-based features were built from demographics, social interactions, and post characteristics. Regression confirms that demographics, mutual relationships, social-media use, the COVID-19 context, and post characteristics each predict knowledge sharing; audience response does not. Among six classification models, tree-based classifiers Random Forest and CatBoost gave comparable AUC-ROC of 0.7805 and 0.7801 respectively; the other four ranged from 0.6689 to 0.7657. Ablation pointed to post characteristics and attitudinal factors as the strongest contributors. Intrinsic motivation and content drive who shares, not how others react.