Studies of interactive regulation and self regulation have tended to exclude each other. However, each person must both monitor the partner and regulate inner state. Thus integration of self and interactive regulation is essential. In this proposal we build on previous work on interactive regulation and add self regulation. In previous work on face-to-face communication in a normal mother-infant sample, a continuum of interactive regulation, ranging from high contingent (vigilant) to low contingent (withdrawn), predicted 1 year attachment outcomes, with midrange optimum. We now hypothesize a midrange optimum in both self- and interactive regulation, and apply these concepts to self-report maternal depressive symptoms (CES-D) and to infant attachment. To disembed the two simultaneous processes, time-series approaches to self- and interactive regulation are used. Data collection is completed: 135 mother-infant dyads have been videotaped and audiotaped in face-to-face interaction at 4 months; 90 have returned for 12 month attachment. Funds are requested for coding and analysis only, using both vocal and kinesic variables. If the current project replicates the predictive power of our previous normal study, our automated vocal instrument will facilitate the detection of interaction disorders at 4 months. If the relative salience of self vs interactive difficulties differs according to kind of pathology (depression/attachment type), translation of this basic research into clinical intervention (currently in progress via videotape coaching) could be more finely focused.
