An automated method of content analysis for psychotherapy research: A further validation

The aim of the study is to validate the ability of ACASM (Automated Co-occurrence Analysis for Semantic Mapping) to provide a representation of the content of the therapeutic exchange that is useful for clinical analysis.

We compared the clinical case analyses of a good outcome psychodynamic therapy performed by a group of clinicians (n = 5) based on the verbatim transcripts (transcript-based analysis) with the clinical case analyses performed by another group of clinicians (n = 5) based on the ACASM representation of the same sessions (ACASM-based analysis). Comparison concerned two levels: the descriptive level and the interpretative level of the clinical case analysis.

Findings showed that, inconsistently with our hypothesis, ACASM-based descriptions of the case obtained worse evaluations than transcript-based descriptions of the case (on all 3 criteria adopted). On the contrary, consistently with our hypothesis, ACASM is undistinguishable from the verbatim transcripts as regards the case interpretation (on 2 out of 3 criteria adopted).

ACASM provides a description of the case that, though different from the one provided by the transcripts, enables clinicians to elaborate clinical interpretations of the case which approximate those produced by clinicians working directly on verbatim transcripts.