5. Homework:

This final section of the training workshop is used to prepare homework. Tow different kinds of tasks are explained and planed in the workshop depending on the objective of the training. When the training is for researchers, the homework is focused in the best procedures for achieve inter-raters reliability. When the training is for clinical purposes, the homework is oriented to use the SOFTA-o for supervision or self-supervision.

a. Live supervision of 3 sessions with the SOFTA-o procedures (for therapist and/or supervisors)

This homework task will be explained and planned at the workshop with supervisors or therapists. Three sessions with different cases will be supervised (when the workshop is for supervisors) or self-supervised (when the workshop is for therapists), focusing on the therapeutic alliance and using the SOFTA-o. Supervisors or therapists will observe each session, marking the behavioural indicators corresponding to each of the four SOFTA dimensions and then rating the scales (from -3 to +3) for each dimension. When the SOFTA is used for supervisors all of these observations and rating scales will be discussed with the supervised therapists after the session and, when possible, a video-recording of the session will be reviewed to give more specific feed-back to the therapist about behaviours indicating a strong or problematic alliance. If possible this homework should be completed during the weeks that follow the workshop. A report about the realization implementation of this supervision or self-supervision task will be sent to the trainers.

b. Rating of 6 video-recorded therapy sessions (for researchers).

Good reliabilities can typically be achieved in 10-15 hours of training. To achieve good inter-rater reliability, we recommend selecting 6-8 practice videotapes that are representative of the investigator's final data set. Ideally, these training tapes should have few clients and good sound and video quality, as well as examples of negative as well as positive Engagement, Emotional Connection, Safety, and Shared Purpose. The difficulty level should be increased gradually as raters become fluent with the task. To assess inter-rater reliability, raters must observe the same sessions independently. During the rating of actual data, members of the rating team should rate the same sessions during the week and come together to compare their results and negotiate to consensus when necessary. Raters should closely compare their behavioural tallies as well as their global dimension ratings. Practice should continue until the raters, as a group, tend to make dimensional ratings that differ by no more than a single scale point at least 90% of the time.