An item response analysis of statistics use in the workplace.

Training in statistics at university should be informed at least in part by what graduates will have to do with acquired statistical knowledge after graduation. A sample of 977 employed graduates with PhD and Masters degrees in seven specialties with statistics pre-requisites at university identifies which of 46 statistics based techniques (the items) they use in their work. A two parameter item response model uses 32 of the 46 items to build a scale measuring the extent of statistics use in the workplace and creates a value for each graduate which is used to summarize differences between the use of statistics in the seven specialties. Implications for syllabus construction to better prepare graduates for the workplace are discussed.