Examines difficulties associated with the use of conversational features in the analysis of mother-infant vocalization episodes. Observations of six mother-infant dyads revealed difficulties involving (1) pooling of data; (2) the unit of analysis; and (3) selection of appropriate "expected" values for dyadic parameters used in tests of…

Classroom research is caught between lack of rigor and lack of richness. When multiple classrooms are used, and when the performance of different groups or levels of students is regarded as "repeated measures" (or split plots) and analyzed accordingly, much of the descriptive richness of the interactions may be regained. (Author)

Suggests that dominance can be defined as asymmetry in predictability in social variables of importance, and bidirectionality as symmetrical predictability. Procedures which address the concepts of cyclicity within a person and synchronicity between people and which assess asymmetry and symmetry in social interaction are discussed. (Author/RH)