This article addresses modeling oscillation in continuous time. It criticizes Steele and Ferrer's article "Latent Differential Equation Modeling of Self-Regulatory and Coregulatory Affective Processes" (2011), particularly the approximate estimation procedure applied. This procedure is the latent version of the local linear approximation procedure based on Boker (2001) and Boker, Neale, and Rausch (2004). It furthermore presents two exact alternative estimation procedures, one using filter techniques and the other using structural equation modeling.