Like other multidimensional alliance measures, the four FTOARS dimensions are not mutually exclusive. Indeed, they are conceptually interdependent. Thus,a family member’s sense of safety in the session is closely related to his or her emotional connection to the therapist and to feeling engaged in the process. The degree to which all family members are active collaborators is reflected in their shared sense of purpose. Reciprocally,when the family shares a common sense of why they are in therapy and what they hope to accomplish, individual family members feel a stronger bond with the therapist.
Interjudge reliability is computed, as an intraclass correlation, for each pair of judges. The final rating, the one used for analysis of the data, should be the consensus rating. We conducted an initial reliability test with 28 archived videotapes of family sessions that had been used in other process studies. Six graduate student judges achieved mean ICC and reliabilities as follows:
Dimensions: ICC Engagement .89 Emotional Connection .88 Safety .74 Shared Sense of Purpose .95
The final rating, the one used for any analysis of the data, should be the consensus rating. That is, if two judges make the same rating (e.g., 0) and the third judge rates the client differently (e.g., -1), the rating used in the analysis would be 0. If the three ratings are all different, the two closest ones could be averaged. For example, the three ratings are -1, 1, and 2. In this case, the -1 would be dropped, and the rating used in the analysis would be 1.5.