This paper describes a novel automated disengagement tracing system (DTS) that detects mind wandering in students using AutoTutor, an Intelligent Tutoring System (ITS) with conversational agents. DTS is based on an unsupervised learning method and thus does not rely on any self-reports of disengagement. We analyzed the reading time and response accuracy of 52 low literacy adults who interacted with AutoTutor to learn reading comprehension strategies. Our results show that students completing a lesson with 20 questions tend to start mind wandering at the 11th ~15th question. Question chunks with mind-wandering have an accuracy of 20%, in contrast to 70% in accuracy for non-mind wandering.