Inter-rater Agreement in Classifying Music According to a Guided Imagery and Music Taxonomy

Guided Imagery and Music (GIM) is a range of therapeutic practices in which clients listen to music selected by a trained practitioner with the aim of gaining cognitive insight through evoked imagery that may be beneficial in working through various inner experiences, pain, or trauma. It is crucial to this process that the chosen music is tailored to the client’s therapeutic goals and receptiveness. Wärja and Bonde (2014) developed a taxonomy consisting of nine categories of musical-psychological characteristics and constructs (e.g., tempo, instrumentation, and mood) aligning with various therapeutic contexts (e.g., supporting and exploring) for helping GIM practitioners select appropriate music, however, its reliability has never before been assessed. In this paper, we present a listening study carried out with 63 GIM therapists and trainees, in order to measure the inter-rater agreement in (1) classifying ten randomly selected pieces from 30 into one or more categories of the Wärja and Bonde (2014) taxonomy, and (2) identifying for each piece heard one or more adjectives from the Hevner mood wheel that best characterize it. Our results indicate participants utilized all categories but with slight to fair overall agreement, however, largely moderate agreement was reported for less musically complex pieces as well as across all pieces when considering only the three primary categories. Our findings support the continued use of the taxonomy and mood for helping select GIM music but suggest also the possible need for clearer descriptions in its subcategories and further training of practitioners who employ it in practice.