As an early member of the Association of Psychiatrists in Training, which was founded to support the changes involved in the new College structure, I would like to comment on the paper by Hughes and colleagues ^{Reference Hughes, Haselgrove, Tovey, Khokhar, Husain and Osman-Hicks1} as well as the issues raised by Shields. ^{Reference Shields2}

Measuring trainees' competence before considering them competent to practise in an unsupervised and leadership role is eminently reasonable. However, it should be understood that such an examination should essentially be a measure of the training programme rather than the individual competence of the trainee. This is not to say that the competence of the trainee is not important, but that the exit examination should not function as a selection criterion. It is simply not fair to allow a trainee who does not have the potential competence to successfully work as a consultant – and this would seem to particularly involve clinical and communication skills – to devote fruitless years to training. For various reasons, candidates considered suitable initially may at a later stage demonstrate problems that preclude a successful transition to a consultant career. Nevertheless, the main selection point should be early enough to allow those regarded as less suitable to pursue alternative areas of medicine for which they may be better fitted.

Thus, while the final examinations should be rigorous, they should focus on more specific areas of knowledge and the expected pass rate of properly trained candidates should be very high. I would also note that the reliability quoted by Burn & Bowie ^{Reference Burn and Bowie3} simply represents a correlation between test items and apart from making the assumption that the results represent interval, rather than ordinal data, says nothing about validity.