Recent research indicates that teachers who share ideas about professional practices are also likely to experiment with educational innovations. Yet, in many Anierican elementary schools, teachers are isolated, each in his own room, precluding effective sharing of resources. This study attempted to induce innovations by encouraging groups of teachers to discuss their professional practices during six zveekly meetings in self-directed T Groups. Memnbers used mimeographed Guide Sheets to take themnselves step by step through exercises designed to (a) develop ties of affiliation, (b) inlcrease knowledge of colleagues inventions, (c) develop norms favoring innovation, and (d) encourage actual use of new ideas. Nine elementary schools (169 teachers) in six suburban districts near New York City cooperated. In each building the faculty was divided into two experimental treatments and a control group. Procedures for both treatments were parallel, but the content was limited to classroom practices for one and to parent-teacher relationships for the other. A check of events during the T-Group meetings showed that teachers learned about colleagues' innovations in five sessions, felt closer to other members during four meetings, and tried some innovations. Norms favoring innovativeness did not develop. Reports of both treatment groups showed that they had tried significantly more innovations than had the controls during the final two weeks of the project. However, in terms of classroom practices, differential effects were found among the buildings, favoring schools where the sessions had produced more knowledge than average about new methods. As expected, once the regular group meetings stopped, all gains in both knowledge of colleagues' new ideas and in experimentation with innovations disappeared. In these schools, weekly self-directed T Groups uapparently functioned with some effectiveness to encourage the utse of innovations by elementary teachers.
