A computer-mediated communication (CMC) system oriented towards changing health-related behavior was developed and evaluated. Stress management training was used to demonstrate the basic technology. Formative research and pilot-testing was conducted, to identify psychological and communication variables that are potentially critical to facilitating behavior change through the CMC medium. The resulting system was used to compare two forms of CMC training with face-to-face intervention: CMC Intensive Intervention (CII) and CMC Non-Intensive Intervention (CNI), with 9 matched subjects per treatment condition.

CII subjects used their personal computers and modems to exchange messages with a therapist (the author), as well as to interact with automated system functions. Such functions included cognitive/behavioral assessment, l recording of self-monitored progress in applying specific and general coping strategies to managing stress, and instantaneous graphic and verbal feedback on such progress.

The CNI form of intervention relied primarily on message exchanges with the therapist, and presentation of general information on stress management. The face-to-face (FFI) treatment was a "traditional" stress management workshop, comprising weekly one-hour sessions over a six-week period. FFI subjects’ assessment, information presentation, self-monitoring, and feedback were analogous to their CII counterparts.

Major findings were that the CII treatment was as effective as the FFI treatment, while the CNI intervention was less effective than the other two treatments, through 3 months follow-up. This finding suggests that active involvement of subjects in practicing specific coping strategies may be critical to efficacious intervention through the CMC medium.

Other findings were that: (1) The CMC message-exchange function can establish a "client-therapist relationship", which potentially overcomes a limitation of "computerized treatment" identified by previous researchers. (2) CMC-based intervention, at least for stress management-related problems, appears to be more cost-effective than face-to-face intervention, for individual treatment but not for group treatment. Implications for extensions to other types of behavior-change intervention and research are discussed.