This article presents the authors' reply to Rod Ellis's comments on their study on motivated classroom behavior. As Ellis correctly summarizes, the three student variables selected for investigation in the authors' study were attention, participation, and volunteering for teacher-fronted activity. These three components were then summed up in a composite measure that the authors labeled "learners' motivated behavior." It seems to the authors that the issue concerning this measure is twofold: (1) Were the constituent variables appropriate to reflect motivated learner engagement in the observed language classrooms; and (2) was the labeling of the variables accurate and appropriate? The authors agree with Ellis that a more finely tuned analysis could have targeted a number of other, more subtle aspects of the students' verbal behavior, which would have also made the three components of their appraisal more distinct. However, theirs was a pioneering, exploratory study, and their main objective at this stage was to achieve a robust and commonsense criterion measure of the students' level of behavioral engagement in instructional events. This is why, although they measured three different aspects of student behavior, they decided not to report any subscores for the constituent components but only focus on their composite score. In the authors' study, "attention" refers to passive academic responding and "participation" to active academic responding. They agree that their labels for these two dimensions were possibly confusing, particularly the term "attention," which inevitably activated a whole spectrum of unintended cognitive meanings. They like Ellis's suggestion of "alertness" and will consider using it in future studies in this vein. In spite of the terminological issue, the authors feel that the composite student engagement measure did work well in actual observational practice. The highly significant correlation with the teacher's motivational practice (r = 0.61) indicated a solid relationship, and it is reassuring that each of the three components displays a significant positive relationship with the criterion measure independently. (Contains 1 table.)