In the preceding 11 chapters, we have reviewed voluminous research and theorizing on dyadic interaction patterns. In light of the extensive work that has already been completed, one might rightfully ask: Is there more to be done? We think so, and so we propose in this final chapter some future research directions that would test our theory and bring the research full circle to connect interaction patterns with outcomes. As a complement to our selective recommendations here, we direct the reader to Montgomery and Duck's (1991) volume, Studying Interpersonal Interaction, which raises a number of additional issues and is full of excellent research suggestions that we will not reiterate here.

TESTING OUR PROPOSED THEORY

Numerous facets of our proposed model invite empirical testing. The model itself requires assessing interactants' required (R), expected (E), and desired (D) behavior levels and partner's actual (A) behavioral level. The RED must be translated into a single Interaction Position (IP) value that can be compared to A so that the degree and direction of the discrepancy between IP and A can be determined. The relative valences of the IP and A must also be assessed. Then the actual predicted adaptation patterns can be tested. We address each of these elements below, followed by some discussion of design and statistical analysis considerations.

Assessing required, expected, desired, and actual behavior levels

The first step in any analysis applying our model is to infer, measure, or manipulate the E, and D elements associated with the communication function at stake.