How is quality of statistics managed and understood for a wider audience?

Abstract

The quality of statistical information is a decisive factor determining the success in fulfilling the mission of official statistics. According to the most common definition, the quality of statistics is assessed by the following seven features: relevance, accuracy, timeliness, coherence, completeness, comparability, accessibility and transparency. This definition puts the users point of view in the first place. It is of crucial importance, therefore, to learn the opinion on information of users as completely as possible at the initial stage of the designing of survey programmes. This paper describes the system of discussions and consultations with users on the annual survey programmes. It presents the users view on the quality of statistical information, and especially the rank given by users to individual features of the quality. Special attention is paid to satisfying the information of users classified as the wider audience. The description of this user group is given, as well as the description of the methods for co-operation with them, aiming at best satisfaction of their expectations on statistics quality. 1. Initial remarks Each producer of goods and services is interested in having products of highest possible quality under conditions of acceptable production costs. This feature constitutes a decisive factor influencing the interest of consumers in a given brand and exerts an influence on their market behaviour. Statisticians must follow similar rules, because the quality of statistical information is crucial for the success in accomplishing the mission of official statistics. It consists of providing the user with high quality information on the economy and society. This information could be necessary for decision-making, for research purposes and for enabling the public to evaluate the efficiency of social and economic development programmes (7). Therefore all NSIs aknowledge the great significance of continuous improvement in the quality of information which they disseminate to users. Attempts in this regard have been particularly strong since the beginning of the 1990s. In many statistical offices we may observe growing awareness of the necessity to adopt a systematic approach to quality issues, which aims not only at improvement of products quality, but also of the whole process of creating and disseminating information, and as a result, - of the quality of the statistical system as a whole (4). Focusing on defining the notion of quality has been a considerable accomplishment of the discussion in the community of statisticians. The starting point for such discussion was a universal definition of quality applied by the International Organisation for Standardisation (ISO) as a totality of characteristics of an entity (a product or service) that bear on its ability to satisfy stated or implied needs (3). In relation to statistics, the most common definition of quality is the one proposed by Eurostat, which assesses quality according to 7 features: relevance, accuracy, timeliness, coherence, completeness, comparability, accessibility and transparency (5). Statisticians of some NSIs prefer to define quality with the aid of a