December 2011

There is so much data collected and stored today, but is this data good data?  Can you be sure that the data you have is the right amount from the right sources to give you the accuracy required to deliver meaningful results?  In this newsletter, Dr John Henstridge explores the idea that planning the data is the first step in any analysis and discusses how statisticians are equipped with the knowledge and tools required to provide assurance in the investment of the data collected.  One of these methodologies is ‘Experimental Design’ which aims to ensure that a data design is appropriately balanced to maximise the efficiency of any analysis.  Dr Kathy Haskard, who recently joined Data Analysis Australia as an experienced Senior Consultant Statistician outlines the process and many benefits of using an ‘Experimental Design’ in the Analytical Ideas Article.  She cleverly uses the AFL football draw as a way of explaining how and why it is used.  Kathy is profiled in this month’s newsletter having had a wide and varied career as a statistician. 

September 2011

Testing for change in a data series over time is the topic of the main article in this month’s newsletter.  Determining change of a known type at a known time is a standard task for statisticians, but often our clients need us to identify sequences of change over time, where the number and timing of these changes is not known.  We discuss two very different examples where Data Analysis Australia has used breakpoint analysis to determine change.  One use was for a quality assurance problem, where laboratory data was tested to check whether results were consistent through periods that had no changes in procedure. The other example was a question of just how much historical data reflected current conditions in order to appropriately identify current operating trends, and to use those trends to produce forecasts.  The Analytical Ideas article looks at a different approach to presenting analysis results.  Though generally results are provided in the body of a written report, recently a number of clients have requested more flexibility.  Data Analysis Australia’s approach was to develop a customised Excel tool to give the clients access to an automated, interactive, user-friendly device, while hiding the often complex statistical methodology.  The staff profile this month is Petra Roberts, who this month celebrates 10 years with Data Analysis Australia.