|Page tools: Print Page Print All|
· content development
· data collection
· data processing
The planning phase of any statistical cycle may be the most important of all phases as it establishes the cornerstone for the collection such as the objectives, timeframes and budget. This phase assumes that investigation of the existence of data to meet the information need has already been undertaken. Concern raised by stakeholders of the NASIS regarding potential duplication of data collection in the system may indicate that sufficient research has not been undertaken. There are existing bodies which aim to minimise duplication and respondent burden in the statistical system, such as the Statistical Clearing House which is the central point for clearance of Australian government business surveys. However, these central clearance points only act for particular streams of statistical collection in the NASIS, which is consistent with stakeholders reporting, for example, that farmers receive similar survey forms from multiple organisations.
This phase also considers the standards and classifications that need to be utilised throughout the statistical cycle to meet the information objectives of the collection. The NASIS is being asked to meet information needs that extend beyond the farm gate and are reflective of the agricultural sector’s supply chain and level of productivity. An accounting framework adopting the principles of the system of national accounts, such as the System of Environmental Economic Accounting for Agriculture (SEEA AGRI) being developed by the United Nation’s Food and Agriculture Organisation, may offer potential for addressing these complex information needs.
The content development phase is largely concerned with the design of the collection including the detailed data requirements, sampling and testing methodology (for a survey). This phase is also an opportunity to consider respondent burden. This is a particular area of concern for stakeholders of the NASR which influences the ability to produce quality statistics. In general, the ABS reports that lower response rates leads to reduced data quality and where a respondent is not engaged in the data collection process (e.g. as a result of burden), the quality of information the respondent provides to a survey is also reduced.
Both ABS and ABARES reported increased difficulty in recent years in collecting data from farmers. Addressing respondent burden is a priority issue in relation to ABS Agricultural Surveys and ABARES Farm Surveys, which were noted by stakeholders as critical agricultural statistical assets. Ways in which data collection could change and result in increased data quality and reduced respondent burden, will be discussed below.
For the data collection phase, a number of innovative suggestions emerged from stakeholders of the NASIS. These suggestions were aimed at increased efficiency within the NASIS, such as the use of online survey forms rather than paper survey forms as a data collection mode. Another area of innovation considers the changing nature of data collection in the future, moving away from survey forms and increasing the use of administrative data sources. For consideration by NASIS stakeholders is how this form of data collection can best be utilised, how it can complement other existing data sources and what kind of enhanced statistical capability is required.
The data processing phase of a collection refers to the processes of coding, checking, editing and applying privacy and confidentiality protection measures to the statistics produced. Suggestions from public submissions referred to the potential for increasing the use of the knowledge and networks of industry bodies by government agencies during this phase. In this, industry expertise could be used in data checking (while maintaining the privacy and confidentiality of data providers) and ensure the production of quality statistics to inform decision making.
The analysis phase of the statistical cycle should be considered as part of the planning phase and includes statistical analysis and interpretation of data, providing insight into the topic being investigated. A statistical collection may involve the collection of information from multiple sources. The transformation phase in this instance would include statistical data integration or the integration of processed information from different administrative and/or survey sources to create a new dataset for statistical purposes. While data integration is not in itself something new, the level of capacity and capability within the NASIS to facilitate this approach may not currently exist.
Statistical data integration would allow users of agricultural statistics to consider multiple dimensions (social, environmental, economic) of the sector simultaneously. While increased use of statistical data integration practices by system stakeholders would increase the effectiveness of the NASIS, its use also opens up challenges in regard to ensuring the privacy and confidentiality of data providers as well as stakeholders needing to adopt common standards and classifications.
The dissemination phase was referred to in multiple ways by stakeholders of the NASR during the public submissions process. This phase revolves around the most appropriate method for delivering statistics and/or statistical analysis to the intended data users. Throughout the NASR’s first phase of consultation, where stakeholders spoke of the need for integration of data from multiple sources, they also described the need for ease of access to these individual and integrated statistical assets. This is an area for potential innovation which could be approached from a number of directions and which the NASR needs to explore further.
Feedback from several stakeholders of the NASIS described the idea of a form of ‘one stop shop’ for statistics informing the agricultural sector with a user-friendly interface that caters for data users with varying degrees of statistical capability. An existing product of best practice, referenced by these stakeholders was the ABS' Census Table Builder. This product sources data from a single collection and applies confidentiality measures 'on the fly'.
The final phase of the statistical cycle, the evaluation phase, is about evaluating the outcome of the processes and considering ways to improve data quality in future. For example, evaluation may reflect that data quality was reduced due to the untimely release of statistics and that for any future collections, improved timeliness is recommended. The level of importance placed on this phase of the statistical cycle was not commented on during the public submission process. However, evaluating the NASIS through contribution to the NASR indicates that stakeholders have reflected on their expectations of data quality and ways to improve the quality of statistics in the system to inform their information needs.
These documents will be presented in a new window.