The use of emotional appeals in political conversations can be an effective strategy to trigger short-term engagement but also result in detrimental societal outcomes as alienation, polarization, and extremism. Whereas scholars examined different individual types of emotional appeals, this study proposes a multifaceted approach to understanding emotional appeals. It applies computational methods to extract tone, discrete emotions and topic modeling to extract topics later classified into appraisal types. Findings show that whereas users primarily posted content that was negative in tone, angry in emotion, and critical in appraisal, Twitter users engaged more with positive, joyful, supportive, or just neutral content.
