Home  >  Sites  >  Helpful Links

Data Sources & Online Analysis e.g. LLM better than ChatGPT

Australian Bureau of Statistics (ABS)

Chris says: The ABS is a treasure trove of reliable information which it gleans not just from the census but other ABS surveys and different government departments. There are many ways to link it with your data such as using demographics or geographic area. If you need support please ask us for help. Ideally before you collect your data as it will need to be collected in ways compatable to linkage e.g. using the same occupation categories as the ABS. This link explains the different data sources and analysis tools e.g. non Census data like the Person Level Integrated Data Asset (PLIDA) - a secure data asset combining information on health, education, government payments, income and taxation, employment, and population demographics with data from a number of government agencies.

StatGPT

Chris says: One of the problems with LLM’s like ChatGPT is that they always want to give you what you ask for, even if it’s wrong. So if you ask for data it will go find it even if that means using unreliable sources. It will report it in proffesional looking tables and visualisations, but the data may be wrong and it’s hard to validate. This is where StatGPT comes in - it only queries respected data sources such as the IMF, Eurostat, World Bank, and ABS. It allows you to query, transform, analyze, visualize, and interpret statistical data using a natural language interface. It will generate code to analyse it, ask pertinant questions to ensure you use it correctly to avoid bias, and highlight any issues with the data.

Our World in Data

Chris says: Another good source of global data. Their goal being to publish the ”research and data to make progress against the world’s largest problems”. Hans Rosling and his Gapminder software were an inspiration, if you’ve never seen his amazing visualisation work check out this short video he did with the BBC Hans Rosling’s 200 Countries, 200 Years, 4 Minutes.

State and Local Governments

Chris says: state and local governments can also be reliable data sources e.g. Open Data Qld.

Ecological Data

Chris says: two good sites for ecological data are the Atlas of Living Australia and Terrestrial Ecosystem Research Network (TERN). There are likely others for marine systems as well.