Abstract

Official statistics in the agricultural sector are used for a wide range of purposes. Many users are satisfied with the summary statistics that are publicized and quality declared in accordance with regulations. However, a substantial number of users have more specific needs. Researchers often require micro data, analysts need statistics at different granularity and structure than those released, and regional users struggle to find statistics that match their regional requirements.

The aim of this study is to examine whether applying the concept of Information Quality (InfoQ) can provide a more thorough understanding of user needs by focusing on the analyses and tasks for which the statistics are used and on how well the data suit these purposes.

A qualitative approach was employed, using semi‑structured interviews guided by the eight InfoQ dimensions, with particular attention to those that discuss the input of data from official statistics in relation to goals and analytical methods. The respondents were also asked to focus on work that had bearing on policy-making.

Four categories of agricultural statistics users were defined: policy analysts, researchers, regional analysts, and areas of the statistical system outside agriculture.

InfoQ, as defined by Kennet and Shmueli (2014) is “the potential of data to achieve a specific (scientific or practical) goal by using a given empirical analysis method.” It is described as a function of the data (X), the analysis method (f), the analysis goal (g), and the utility of the results (U). The concept is broken down into eight dimensions to assess quality. This method has previously been applied to official statistics (Kennet and Shmueli (2016)

Preliminary results show that the InfoQ framework provides a more user oriented perspective and is useful for evaluating how well official agricultural statistics meet the diverse analytical needs of users, thereby offering guidance for improving data granularity, structure, and accessibility. Nevertheless, the focus on the analytical methods and specific goals of different studies also make the result difficult to imply for developing the official statistics. The qualitative work could also serve as a starting point for more quantitative studies.

The preliminary findings indicate that analysts also use research like analytical methods that require micro data, and they must adopt other methods or outsource projects to fulfil some of their needs. Another dimension for analysts is temporal relevance, specifically the time lag between the reference period and data availability.

For researchers, the underlying construct is often more interesting than the data itself, but it is essential that the dataset includes the specific variables needed for the analyses. At the regional level, data structure and granularity are among the main challenges

The broader statistical system typically has specific data needs for feeding models used in further applications. In these cases, granularity and data reliability are important. These needs are often long term, whereas analysts and researchers frequently have varying needs depending on the policy context.

1 Kenett, R. S., & Shmueli, G. (2014). On information quality. Journal of the Royal Statistical Society. Series A: Statistics in Society, 177(1), 3–38. https://doi.org/https://doi.org/10.1111/rssa.12007

2 Kenett, R. S., & Shmueli, G. (2016). From quality to information quality in official statistics. Journal of Official Statistics, 32(4), 867–885. https://doi.org/10.1515/jos-2016-0045