The 2009 Australian and New Zealand Stata Users Group meeting has proven to be a great achievement, gathering people from different disciplines and across Australia and New Zealand.
The first Australian and New Zealand Stata Users Group (ANZSUG) meeting was a big success when it was held in Adelaide in 2004. The second was held in Melbourne in September 2006. This third meeting was convened by MEAFA and held at The University of Sydney from the 4th to the 6th of November, 2009. The attendance at the Stata Users Group meeting rose above our initial predictions, with almost one hundred people subscribing for the event.
MEAFA was overwhelmed with the background diversity of the participants, what proportioned fruitful discussions and the unique atmosphere to exchange experiences among Stata users. The participants came not only from academic institutions all over Australia and New Zealand, but also from the regulatory body such as the Reserve Bank, Cancer Council NSW, Food Standards Australia New Zealand (FSANZ), Australian Institute of Family Studies, The George Institute for International Health, and many others. The private companies also marked their presence with attendants from Frontier Economics, Queensland Health, St Vincent Hospital, The Queen Elizabeth Hospital, etc.
The forum comprised structured thematic sessions covering an eclectic range of topics related to the usage of Stata for analysis of biostatistics, econometric, psychological data, etc. Stata developers from the United States hosted a “Wishes and grumbles” session, given by Roberto G. Gutierrez, StataCorp’s director of statistics, and Alan Riley, StataCorp’s vice president of software development. In addition to the meeting two workshops were organized: Introduction to Stata programming and Mata matters, on November 6, and Analyzing complex survey data using Stata, on November 4.
The aim was to bring together a wide range of users to discuss statistical analysis, data management, and graphics using Stata. Speakers were encouraged to present their experiences with the package, whether using standard commands, writing new programs, or exploiting the graphics facilities.
Ron Bose, from International Initiative for Impact Evaluation, started the presentations talking about the impact of water supply and sanitation interventions on child health, with evidence from Demographic and Health Surveys (DHS). He examined the impacts on child health with access to different types of water and sanitation facilities, and from socioeconomic and child specific factors. Using multiyear cross-sectional health DHS data, he employed the quasi-experimental estimators (matching) to match children belonging to different treatment groups, defined by water types and sanitation facilities, with children in a control group. Quantile regression models were used to benchmark results and to check for their robustness. The estimates served as an input into cost-effectiveness analysis that compares the provision of increased access to sanitation with other public health interventions in developing nations (especially sub-Saharan Africa and South Asia) to underline its importance in achieving Millennium Development Goals (MDGs).
Maya Guest, from The University of New Castle, targeted the use of a quantile regression model in an occupational health and safety study. She presented the hearing threshold data collected for the SHOAMP study to compare the hearing thresholds with the reference population using quantile regression. The model estimated median hearing thresholds significantly lower than normal and the extent of the hearing loss is substantial in that a 95% confidence band for the median lies below the 30th percentile of the normal population for most frequencies and ages. She reported that largest loss occurs at 6 and 8 kHz for those employees with less than 30 years of age.
Peter Herbison, from University of Otago, discussed the effects from the lack of independence in meta-epidemiology. He pointed out that papers submitted from this kind of data have been heavily criticized because of the feared effects of lack of independence and suggested that one meta-analysis should be randomly select from each systematic review for the sensitivity analysis. As an extension to that he chose to bootstrap the results of interest to examine the effects of the lack of independence. The results of two such sensitivity analyses showed that the lack of independence appears to have little effect on the interpretation of the results.
Ahmad Rabiee and Ian Lean, from SBScibus, showed in their presentation a meta-analysis in animal health and reproduction, with methods and applications using Stata. They used many of the statistical methods to conduct meta-analysis and demonstrated how Stata can provide a comprehensive suite of programs that can be used in meta-analysis. Details on the common statistical methods used were presented and examples of when these have been used in studies using cattle were provided.
Kieran McCaul, from University of Western Australia, explored the issue of generating RTF files in Stata to create tables for inclusion in Word documents. He argued that when users write manuscripts in Word, a table created as an RTF file can be included into a Word document quite easily, and generating such a table from within a Stata program requires knowledge of only a few RTF commands. While this is a “brute force” method for generating a table, it is particularly useful if many tables of the same layout have to be generated (in a thesis, for example), or where an annual report has to be produced and the number and layout of tables remains essentially the same from year to year. He described the RTF instructions required to produce a table and outlined the approach that he took when generating an RTF file from within a Stata program.
Karl Keesman, from Survey Design and Analysis, talked about automating reports with Mata and mail merge, and suggested that Stata and Mata can generate an output that mail merge can read. The commands and process of doing this with Mata were explained and demonstrated using a simple example.
Philip S. Morrison, from Victoria University of Wellington-New Zealand, analysed the issue of complementing Stata with geovisualization. He explained that Statistical agencies are increasingly recognizing the value of configuring their data in formats that facilitate geovisualization. For Stata users this poses a challenge because the present geovisualization capacity within the conventional Stata product is quite limited. His presentation reported on a project completed for Statistics New Zealand on Geovisualization where graphical tools from Stata were complemented by the geovisualization capacity afforded by GeoViz. The presentation illustrated the returns to geovisualization via a specific case study and considers the advantages that could potentially accrue to Stata users if such a capacity was provided within the Stata system itself.
Jisheng Cui, from Deakin University, offered a more applied view of working correlation structure and model selection in GEE analyses of longitudinal data. The work was based on the QIC method proposed by Pan (2001, Biometrics 57: 120–125), and systematically developed a general computing program to calculate the QIC value for a range of different distributions, link functions, and correlation structures. The QIC value can be used to select both the best correlation structure and the best subset of explanatory variables. The presentation introduced the QIC method and program written in Stata software, and demonstrated how to use it to select the most parsimonious model in GEE analyses through several representative examples.
More details are available at Stata User Group webpage.