7105.0.55.004 - National Agricultural Statistics Review - Final Report, 2015  
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 29/07/2015  First Issue
   Page tools: Print Print Page Print all pages in this productPrint All

3. USE OF BEST PRACTICE METHODS AND SOURCES TO MAXIMISE DATA QUALITY AND MINIMISE RESPONDENT BURDEN

Best practice agricultural statistical systems strive to develop statistical assets and supporting infrastructure to a quality standard that is fit-for-purpose for the information needs they are intended to inform. Around the world, national statistical offices have adopted quality frameworks that are used to assess and report on the quality of statistical data, and to assist those developing statistical collections to produce high quality outputs. The ABS Data Quality Framework is one such framework53, as is the FAO Statistics Quality Assurance Framework 54. These frameworks recognise that quality in statistics has a number of dimensions, including relevance, accuracy, timeliness, accessibility, coherence, and interpretability, and is a function of a range of elements, including the institutional environment within which the statistics are produced, as well as the statistical practices and processes used.
The FAO notes that quality depends on user perspectives, needs and priorities, which vary across groups of users:

      For this reason the major challenge is to achieve a compromise among the needs of the various possible users (current and potential) in order to produce and disseminate statistical outputs that satisfy the most important needs given constraints concerning available resources55.
Effective engagement with users in the statistical system is required to understand and prioritise quality requirements, and engagement with policy-makers and funding bodies is required to ensure the statistical system is sufficiently resourced to meet these requirements.
Producers of statistics should also ensure that they choose the most appropriate method of developing or compiling the required statistics. A fundamental principle of best practice statistical production, accepted internationally, is that sources and methods of data collection should be chosen to ensure timeliness and other aspects of quality, to be cost-efficient and to minimise the reporting burden for data providers56. Best practice statistical organisations leverage a range of data sources to produce their statistics, recognising that each type of source (e.g. directly collected from respondents via surveys, or from administrative or other sources) has strengths and weaknesses and that sources can be combined to produce a better product. The US Department of Agriculture, for example, draws on a mixture of administrative and survey data, remote sensing and field observations to produce its statistics.
A challenge for producers of statistics is balancing the need to meet users’ statistical requirements with the need to effectively manage the burden on respondents. Internationally, the importance of effectively improving the management of the relationship with survey respondents has been recognised by a number of statistical organisations, with agencies such as Statistics New Zealand and the Office of National Statistics in the UK implementing reviews and improvements to their engagement with survey respondents. Statistics New Zealand's Respondent Experience Strategy 2013-2057 explicitly recognises that respondent engagement involves more than managing survey load, and that improving the respondent experience58 is fundamental to significantly addressing the reasons behind the reluctance of respondents to comply with survey requests. The Strategy aims to achieve 'willing respondents, finding it easy to comply' through putting respondents at the centre of data collection. The Strategy involves: increased levels of advocacy for respondents' rights and interests; more rigorously balancing compliance against the benefits of what is collected; taking a more fit-for-purpose approach to compliance; working to 'get people on board' with compelling reasons and benefits for survey participation; and 'making it easy' to comply by either reducing the amount of data being asked for or by making it easier to provide that information.

FOOTNOTES

53 The ABS Data Quality Framework is based on the Statistics Canada Quality Assurance Framework (2002) and the European Statistics Code of Practice (2005) – see ABS 2009, ABS Data Quality Framework, May 2009, cat. no. 1520.0, ABS, Canberra.
54 United Nations Food and Agriculture Organisation (FAO) 2014, The FAO Statistics Quality Assurance Framework, FAO, Rome.
55 Ibid, p.6.
56 Principle 5 of the United Nations Statistics Division’s Principles Governing International Statistical Activities, endorsed in 2005. A similar theme is found in Principle 5 of the Fundamental Principles of Official Statistics, endorsed by the UN Statistics Commission in 1994 and Principle 14 of the ABS-developed National Statistical Service (NSS) Key Principles
57 New Zealand Government's Statistics New Zealand's Respondent Experience Strategy for 2013-20, Wellington: Statistics New Zealand.
58 Statistics New Zealand defines respondent experience as follows: "Respondent experience refers to the perceptions, feelings and reactions that a person has a result of completing our surveys. A respondent's experience is important as it affects their willingness to comply with our requests for data and also to provide complete and accurate information. This, in turn, impacts our ability to achieve our response rate and data quality targets and produce statistics efficiently." - source: New Zealand Government's Statistics New Zealand's Respondent Experience Strategy for 2013-20, Wellington: Statistics New Zealand, p. 1