This research is based in Australia. To what extent can its findings be applied to other countries where substantial numbers of prospective employees, often migrant or poorly educated workers, do not have the language in which selection tests are constructed as a first language? What are the implications of this?