1001.0 - Australian Bureau of Statistics -- Annual Report, 2013-14  
Previous ISSUE Released at 11:30 AM (CANBERRA TIME) 04/11/2014   
   Page tools: Print Print Page RSS Feed RSS Bookmark and Share Search this Product

METHODOLOGY AND DATA MANAGEMENT

ANALYTICAL SERVICES

The branch conducts R&D in data access, integration, confidentialisation and analysis methods for traditional and emerging data sources. It also supports business-as-usual (BAU) statistical production in the ABS through the provision of analytical products and methodological advice to subject matter areas. As part of its BAU support function, Analytical Services delivers specialist consultancy services to statistical users in other government agencies.

By providing analytical products and methodological advice, the branch contributes to the delivery and continued improvement of ABS statistical outputs. In particular, time series in seasonally adjusted and trend form are an important input to policy formation, decision making and research in government, academia and the private sector. The advancement of data access, integration and analysis methodologies underpins the creation of a richer, more dynamic and focused statistical picture of Australia for better informed decision making by government, business, academia and the Australian public. Dynamic data confidentialisation methods, especially those applied to microdata, are intended to provide robust protection for the confidentiality of information provided by individuals or businesses.

Key achievements in 2013–14

  • Advanced ABS capability in micro-simulation was achieved through joint work with Treasury on CAPITA, a new general purpose static tax-transfer micro-simulation model based on ABS data.
  • New demographic methods for understanding inter-censal Indigenous and Torres Islander population estimates and improving future estimation were developed.
  • New insights were delivered through econometric and statistical modelling of cross-sectional and longitudinal data—changes in individual labour force status over time; relationship between innovation and culture for productivity; household energy consumption patterns; and the relationship between science, technology, engineering and maths (STEM) skills, R&D expenditure, innovation and collaboration.
  • Leading-edge disclosure avoidance techniques in TableBuilder and Data Analyser were developed and implemented, and improved methods and tools for confidentialising business survey data were evaluated.
  • New methods and an innovative information platform for using linked employer–employee data in firm-level productivity analysis were developed, and new linked statistical datasets—the ACLD and Census–Settlements linked dataset—were created.
  • Big data strategy and framework were developed and selected big data applications for official statistics were progressed.
  • Work contributed to international projects that are developing common components for data linking, confidentiality, seasonal adjustment, and data analysis for official statistics.

DATA STANDARDS AND METHODS

The branch is responsible for promoting the comparability, integration and quality of ABS statistics, through the use of standard concepts, definitions, classifications and procedures. It is also responsible for the infrastructure used to hold key definitional metadata and to store statistical data from which ABS outputs are sourced. A key focus of the branch is the Statistical Metadata Transformation Program which will plan for the transformation of metadata into future ABS 2017 infrastructure

Key achievements in 2013–14

  • Coherence of statistical metadata was improved through the promotion of corporate metadata.
  • More international-based standards and classifications were utilised.
  • Four publications relating to updated classifications and standards were released.
  • Methodology architecture was developed to steer the development/adoption of statistical methods and tools to support ABS 2017.
  • Governance framework and strategies were developed for managing ABS metadata, including concepts and classifications.
  • ABS responses to the UNSD COICOP (United Nations Statistics Division Classification of Individual Consumption According to Purpose) review were co-ordinated and made.

STATISTICAL SERVICES

The branch provides specialist services to meet new and ongoing demands in official statistics. Much of the work has the goal of ensuring the methods underlying ABS outputs are based on sound, defensible statistical principles and are cost effective.

The branch has specific responsibilities for supporting statistical collection and production processes. It provides advice on and methods for the most efficient and effective ways of meeting information needs; on data quality, through all stages of the survey cycle; and undertakes research on statistical and operational research methods to improve the efficiency and reliability of ABS work.

The branch also aims to minimise the load on data providers through efficient data collection methods, sample designs, techniques to control sample overlap between surveys and, in the case of Australian business surveys, through reviews and approvals by the Australian Government Statistical Clearing House which is managed by the branch.

Key achievements in 2013–14
  • Methodology for design of the new Freight Movement Survey was developed and integrated with the Survey of Motor Vehicle Usage.
  • Standards were developed and assistance provided for the ABS implementation of web forms for business surveys and the Labour Force Survey.
  • It was demonstrated that response rates could be lowered for the Labour Force Survey with minimal effects on the quality of the statistical outputs, thus delivering substantial savings in collection costs.
  • The measurement of statistical impact from introducing web forms to the Labour Force Survey was concluded.
  • Enhanced respondent engagement techniques in economic, labour and social surveys were developed and deployed.
  • Unnecessary provider load was reduced or eliminated through effective implementation of the Statistical Clearing House policy.
  • A modelling framework was developed to predict the number of field staff needed for the follow-up phase of the 2016 Census.
  • A two-phase weighting methodology for dealing with missing data in a linked dataset was developed and evaluated.
  • An inventory of ‘Usage of administrative data sources for statistical purposes’ was contributed to the international group on Methodologies for an Integrated Use of Administrative Data (MIAD) in the Statistical Process.