Australian Statistical Conference 2004

Statistics is not usually a very visible profession, so the sight of 780 statisticians in Cairns was impressive. They were the attendees at two parallel meetings - the 2004 Australian Statistical Conference and the International Biometric Conference. This produced five days of intense seminars, workshops and, as at any conference, innumerable discussions as professionals bounced ideas, new techniques, and problems off each other.

These conferences highlighted one of the most important features of statistics - diversity. Those who attended the conferences had job titles such as biometricians, statistical geneticists, analysts, medical statisticians and research officers as well as those who go by the title of statistician. Attendees worked for universities, pharmaceutical companies, hospitals, research institutes and consulting groups. Some were highly specialised in a particular area of statistics while others applied their statistics in a more diverse set of areas.

The majority of those present at the conferences gave presentations and they reflected this diversity. Some examples are, Responding to Bioterrorism: Models for Anthrax, Mixed Maggots: The Analysis of a Fly Competition Experiment, Statistical Methodology for Facial Identification, Malnutrition in Nigerian Children (0 - 5 years): Distribution and Classification, and An Investigation of Gambling in Cherbourg.

The talks presented by the Data Analysis Australia contingent were no exception to the display of diversity. Meredith Regan presented the issues of modelling water consumption and assessing the effect of water restrictions on overall consumption. Anna Munday discussed the modelling of internal migration patterns in the context of population forecasts. And John Henstridge spoke about the sampling design that Data Analysis Australia has used when conducting the Perth and Regions Travel Survey (PARTS).

Data Analysis Australia also had a strong presence at the Young Statisticians session, being organised by Anna Munday and chaired by Jodie Thompson, with John Henstridge as one of the invited speakers. One of the issues discussed at this session was the need to attract more students to study statistics. It was generally felt that there was a need to convey to high school students the diversity of career opportunities that statistics offers in order to boost future numbers studying statistics.

One of the highlights was the statistical consulting session where the invited speaker was Stephen Fienberg from Carnegie Mellon University in Pittsburgh. Stephen presented a talk with the intriguing title In search of the magic lasso: The truth about the polygraph. (The title of the talk came from the fact that the inventor of the lie detector went on to create the Wonder Woman comic!). Stephen related his experiences of chairing a high level committee of the National Research Council that was asked by the United States Congress to report on the utility of the lie detector for screening employees of secure government laboratories.

The main finding was that the polygraph had such an error rate that any attempt to use it to catch spies would be swamped with false positives (loyal employees who would be incorrectly classified as suspects) and many of the technical statistical issues were covered. But the most interesting and relevant part of the presentation was the method used to convey technical information to a non-technical audience. In the end, the report used a simple numerical example - a pair of small tables - that even a politician could get right. Another example of the diversity of statistics!!

October 2004