Reliability of coding affective communication in family therapy sessions: Problems of measurement and interpretation.

Abstract

Coders have difficulty in achieving a level of 70% agreement in coding transcripts of affective interaction in conjoint family therapy. Contributing factors are (a) difficulties in defining coding categories and coding rules, (6) differences in coders' sophistication, and (c) variations in their attitudes toward coding. These difficulties have been partially solved by defining the coding categories and rules more rigorously. Coders may then still disagree on who is speaking to whom because of ambiguity in intrafamilial communication. When coders disagree on who is speaking to whom, they also disagree on the type of affect being expressed. If family members have the same difficulty as coders, therapists should focus on helping the family clarify indirect communication, thereby improving affective problem solving.Many authors have documented the process of reliably coding individual psychotherapy sessions (

Process and Outcome in Conjoint Family Therapy

“…They learned the coding system during a training period using a manual containing detailed coding rules (10). This manual was the final outcome of several earlier versions that were revised as a result of studies of inter‐judge agreement (11). Using this manual, coders can achieve an overall level of agreement of seventy‐five per cent.…”

“…When the actual process was studied by Sigal, Rakoff, and Epstein (21), the report was based on material recalled by the therapist rather than material derived from direct observation. No doubt one of the reasons for the lack of studies involving direct observation has been the considerable methodological problems involved in recording and measuring family interaction (8, 11, 22). Bales (2), Dreschler and Shapiro (6), Haley (12), Riskin (18), and Terrill and Terrill (24) have reported methods of describing and categorizing group interaction, but their application so far has been primarily in situations of group discussion or family problem‐solving tasks.…”

The McMaster Model of Family Functioning

“…An important aspect of this conceptual model is its clinical utility. It has been developed and used extensively in a variety of psychiatric and family practice clinics (Comley, 1973; Epstein & Westley, 1959; Guttman, et al, 1971Guttman, et al, , 1972; Postner, et al, 1971; ; Sigal, et al, 1967; Westley & Epstein, 1960) and by therapists who treated families as part of a large Family Therapy Outcome Study ; Santa-Barbara, et al, 1975; Woodward, et al, 1974Woodward, et al, , 1975Woodward, et al, , 1977. The framework has also been used in a Family Therapy Training Program' and found to be readily teachable.…”

The relationship between therapist behaviors and outcome for middle- and lower-class couples in marital therapy

“…Hollis (1968a), who used a verbal interaction typology developed earlier (Hollis, 1967(Hollis, , 1968b, found that caseworkers who used more reflective comments instead of directive and ventilative comments in the initial joint couple interviews had lower rates of discontinuance. In a similar vein, Postner, Buttman, , who used a process coding scheme developed at the Jewish General Hospital in Montreal for studying verbal interaction in family therapy (Guttman, Sigal, & Chagoya, 1972;Guttman, Spector, Sigal, Rakoff, & Epstein, 1971;Sigal, Guttman, Chagoya, & Lasry, 1973;Sigal, Rakoff, & Epstein, 1967;Spector, Guttman, Sigal, Rakoff, & Epstein, 1970), examined the effects of therapist directive (D) statements (stimulating interaction, information gathering, giving support) and interpretation (I) statements (clarifying motivation, labeling unconscious motivation). The DirectiveInterpretation ratio decreased as therapy progressed.…”