Abstract
Student counsellors' pre-test, post-test and retrospective pre-test scores on the Counselor Evaluation Rating Scale (CERS) were used to investigate the presence of a ‘response-shift bias’, to estimate the overall effectiveness of training and to promote counsellors' continuing self-reflection on their work. Results from three groups of trainees spanning 3 years revealed the presence of a response-shift bias, confirmed the effectiveness of the training programme and demonstrated the students' continuing self-analysis of their own counselling skill.