The results should have been cross-validated using a holdout sample to compensate partially for the use of methodologies that capitalised on change. Unfortunately, the psychologist’s admonition in his report to cross-validate the obtained results before implementing the selection procedure was under the described circumstances unlikely to happen. The client presumably had the expectation of obtaining a usable selection procedure upon the completion of the project. A requirement for more data (which, because of the low rate of hiring, would not have become available in any reasonable period of time) before the selection procedure could be released for use was technically correct but practically unlikely to happen.