The First-Grade Reading Group Study is an experimental examination of teaching behaviors and their effects in first-grade reading groups. The specific teaching behaviors of interest are defined by a model for small group instruction which describes organization and management of the class, and ways of responding to children's answers that are associated with achievement gains. The coding system was designed specifically for this study, although several categories were adapted from the Brophy-Good Dyadic Interaction Coding System. This system is to be used only in reading groups and only to measure public encounters between the teacher and individual members of the class. There are two kinds of forms used, a high-inference and a low-inference coding sheet; both are appended. On the high-inference sheet are check lists and rating scales describing various events which do not involve interactions between a teacher and a single child within the context of the reading group, i.e., teaching techniques, student behavior, or seating arrangement. The low-inference sheet is used to describe each step in an interaction between the teacher and an individual child within the context of the reading group. This involves both academic and behavioral interactions. (Author/MV)