Projects Current Page Projects


Computer-aided exercises in psychomotor diagnosis

Background and aims

The skills of psychomotor therapists exert a major influence on the quality of psychomotor therapies. This is particularly so with diagnostics, a core skill in the educational therapy professions. The quality of diagnostics largely determines how effective a therapy will be. Experience with specific individuals and situations plays a key part in building diagnostic skills. When studying, this experience is usually acquired from supervised work experience and the accompanying sessions where students are able to reflect on their experience. The proposed study seeks to reinforce the acquisition of diagnostic skills while studying for a degree. Ideally the solution will require hardly any additional human resources.


How can greater room be given to diagnostic experience, combined with informed feedback, in the training of a psychomotor therapist? Can it be achieved with minimum human resources? The proposed study seeks to implement these requirements by using computational linguistic procedures and video recordings. What form must a training programme take so that students can practice as intensively as possible and continually review their progress with the aid of feedback – in the spirit of deliberate practice (Ericsson, Krampe & Tesch-Romer, 1993)?


A computer-aided learning environment will be developed in partnership with the Institute of Computational Linguistics at the University of Zurich. It will enable students to watch extracts from a specific psychomotor case evaluation, to begin recording their ideas for diagnosis and therapy online and to improve these stage by stage by repeatedly retrieving feedback about how closely their own texts tally with texts about the same videos written by experts and stored in the system.

By using so-called distributional semantic methods, it should be possible to describe the similarity between freely formulated texts and to feed the results back to the trainee. A similar method has been applied by the Department of Psychology at the University of Bern in the group working with Franz Caspar and has been assessed as effective (e.g. Caspar, Berger & Hautle, 2004).

The proposed project breaks down into the generation of a video file, the construction of a “semantic space” as a basis for text comparisons, the generation and integration of diagnostic expert texts, and finally the evaluation by advanced students of a first version of the training programme.

The methodology for this pilot project centres on technicalities of method and practical didactics. A proposed follow-up study aims to review the efficacy and ecological validity of the procedure in greater detail.


An analysis program that can compare students’ texts with experts’ texts and report their similarity was created in accordance with the aims, although with time-delayed feedback instead of online feedback as had been planned. The evaluation of the system through comparison with similarity estimates made by people confirmed the potential of the method as regards a follow-up study. However, it also suggests technical improvements in view of practical application.


  • Caspar, F., Berger, T. & Hautle, I. (2004). The Right View of Your Patient: A Computer-Assisted, Individualized Module for Psychotherapy Training. Psychotherapy: Theory, Research, Practice, Training, 41 (2), 125-135.
  • Ericsson, K. A., Krampe, R. T. & Tesch-Romer, C. (1993). The Role of Deliberate Practice in the Acquisition of Expert Performance. Psychological Review, 100 (3), 363-406.



Project Management

Dr. phil.  Müller



Forschung und Entwicklung
Tel: +41 44 317 11 81

zfe[at] zfe