The Quality of Violence Risk Formulation and a Quasi-Forensic Mental Health Training Experiment

Abstract

The study was aimed at measuring the impact of training on forensic case formulation in mental health and to provide more evidence on the reliability of the TEC-F. Nine psychiatrists and six psychologists from various Latin American countries participated in a quasi-pedagogical experiment. The quality of formulations was independently and blindly measured pre- and post-intervention with the TEC-F and the assignment of two standard vignettes was also randomly manipulated. Quality mean differences and instrument reliability indicators were calculated. The values of intraclass correlation coefficients were 0.92; 0.94; 0.83; 0.93 and 0.95 and the values of Cronbach's alpha coefficient were 0.83; 0.94; 0.63; 0.77 and 0.93, for the dimensions transparency, specificity, communication, reasoning and for the total TEC-F respectively. The results of the 19-day test-retest were excellent. The mean TEC-F total quality pre-course was 31.4 and the mean post-course, 38.4 (p = 0.003 and p = 0.001 for group test and paired test respectively). The pedagogical intervention produced a significant improvement in the quality of the expert’s formulations. The study added evidence supporting the TEC-F reliability.