An estimation of a nonlinear dynamic process using Latent Class extended Mixed Models: Affect profiles after terrorist attacks

Conceptualizing affect as a complex nonlinear dynamic process, we used Latent Class extended Mixed Models (LCMM) to understand whether there were unobserved groupings in a dataset including longitudinal measures. Our aim was to identify affect profiles over time in people vicariously exposed to terrorism, studying their relations with personality traits. The participants were 193 university students who completed online measures of affect during the seven days following two terrorist attacks (Paris, November 13, 2015; Brussels, March 22, 2016); Big Five personality traits; and antecedents of affect. After selecting students whose negative affect was influenced by the two attacks (33%), we analysed the data with the lcmm-package of R. We identified two affect profiles, characterized by different trends over time: The first profile comprised students with lower positive affect and higher negative affect compared to the second profile. Concerning personality traits, conscientiousness was lower for the first profile rather than for the second profile, and vice versa for neuroticism. Findings are discussed for both their theoretical and applied relevance.