Generating hypotheses about the function of student problem behavior by observing teacher behavior

We examined whether, as predicted by research on child effects, we could generate hypotheses about the function of student problem behavior by observing the amount of attention teachers provided to students. In the first phase of the study, we observed the amount of attention teachers distributed among small groups of students who exhibited problem behavior in individual or small-group instructional settings (problem behavior presumably maintained by attention or escape). Based on the amount of attention each student received, we generated hypotheses about the function of his or her problem behavior. In the second phase of the study, we determined the accuracy of these predictions by conducting a brief functional assessment with each student. Results confirmed that, for 14 of the 15 students, we were able to generate accurate hypotheses about the function of their problem behavior. These results suggest the potential efficacy of using the amount of attention teachers distribute among groups of students to generate empirically based hypotheses about the function of student problem behavior maintained by attention and/or escape. These results also illustrate the efficiency of this procedure; by observing teacher behavior, we were able to generate hypotheses about the function of problem behavior for several students at one time.