The role of decision making in the negative impact of content moderation.

Abstract Content moderators view hundreds of negative images daily to remove content that violates social media guidelines (e.g., live suicides, child exploitation). Frequent exposure to such content can lead to depression, anxiety, and posttraumatic stress disorder. Therefore, it is important to investigate factors that exacerbate the negative impact of viewing graphic content. Here, we examined whether the types of decisions moderators make (i.e., those requiring conceptual processing of content meaning) influence content-related intrusions, state anxiety, and negative affect. Across three experiments, we randomly allocated participants to one of two conditions: in Experiment 1 ( N = 222) and Experiment 2 ( N = 225), participants either made decisions about image content (conceptual processing; experimental ) or about image composition ( control ). Experiment 3 ( N = 221) replaced the composition control condition with a no-task control where participants only viewed the images. In all experiments, participants recorded and rated any image-related intrusions they experienced and completed pre- and post-task measures of state anxiety and negative affect. In Experiment 1, participants who made content decisions experienced fewer intrusions and rated them lower on unwantedness and emotional intensity than those in the composition condition. However, Experiment 2 revealed no differences between these conditions, and Experiment 3 only showed that participants who passively viewed images experienced a greater increase in negative affect compared to participants who made content decisions. Our findings suggest that content-related decisions do not contribute to content moderation’s negative impact. Future research should explore other aspects of content moderation that may exacerbate negative outcomes.