"A recent article in this journal provided a critique of the design of the Survey of Employees Questionnaire within the 2004 Workplace Employment Relations Survey. The principal criticisms concerned the use of vague response categories, double-barrelled questions, needless ordinal measurements and the use of multiple binary items in place of ordinal or scalar measures. The critique highlighted some areas worthy of attention but we argue that, in other areas, it took insufficient account of the survey's core objectives and of the practical and financial constraints guiding its design. We offer a broader assessment of sources of error within the WERS 2004 Survey of Employees. "

Support: Paper