First, a new measure of MI for reaction times is proposed that takes the entire RT distribution into account. Second, we present some recent developments in TWIN modeling, including a new proposal for the sound-induced flash illusion (SIFI).