Chair: Krzysztof Jajuga
Room: S3A Barbakan
Time: 14:30 - 16:00
Date: 27 June
|Title: <<< Communicating official statistics and it´s quality in blogs - a rhetorical analysis >>>
The use of social media as an additional channel for communicating official statistics is increasing. Social media provides possibilities of highlighting and explaining statistics and builds the ethos of NSI:s for new groups of users. At the same time, texts in social media have a tendency to be oversimplified and overstated. I.e. social media places new demands on the NSI:s. In this article, rhetorical theory is used to analyse how the quality of official statistics is described in social media compared with how it is described in publications and databases. The case of how the Swedish Board of Agriculture, the organisation responsible for official agricultural statistics in Sweden, describes the quality of official statistics on consumption of food is chosen. A blog, a traditional publication and a database are compared. First, differences in the rhetorical situation, i.e. audiences, constraints and exigencies between the three ways of presenting statistics are highlighted. According to rhetorical theory ethos, pathos and logos are essential for building trust and the article therefore discusses how trust in the statistics might be affected by the use of social media as opposed to the more traditional channels for communicating official statistics. The findings show that the blog focuses on the consumption of meat, i.e. a type of food that is widely discussed in media. The blogposts are used by media and the public and are timely posted when meat consumption is debated in the society. The language in the blog is simplified compared with the language in the publication. The description of quality is mostly concerned with one aspect content. While the description of the quality was not as thorough in the blog as in the traditional publication, it was easier to comprehend and covered the aspects most essential to the content presented.
|Title: <<< Collaboration with universities – a way to promote innovation and experimental culture at statistical offices >>>
Statistics Finland has had fruitful collaboration with the University of Helsinki and the University of Jyväskylä for over several decades. Thesis traineeships are an important part of this collaboration. Each year, a handful of students are recruited for a period of six months. The main task of the student is to write a thesis on a given topic. Statistics Finland provides the student with a research problem, data set, facilities and day-to-day guidance. The academic supervisor is from the student’s university. This paper presents how a research problem is drafted in collaboration with the academia, how the students are recruited and supported during the traineeship as well as what the trainee program has resulted. We illustrate by concrete examples how the thesis traineeship can promote innovation and experimental culture in a statistical office. These examples involve testing new tools and methods. We argue, that the thesis traineeship is a good way to involve the scientific community in the activities of a statistical authority. In particular, we show that the thesis traineeship is not only a mutually beneficial practice for the academia and statistical offices but also for the students.
|Title: <<< How widening access to researchers has improved and can continue to improve microdata quality ? From national experiences to European aspects. >>>
There are now a number of experiences at national level where opening access to the official microdata for researchers has been an important driver for improving microdata quality both for the metadata and for the data themselves which are two closely associated faces of data quality. Based on the French experience, the paper provides examples and discuss possible developments with increasing access to confidential microdata both at national and European level. Regarding metadata quality, one important aspect of the GSBPM (Generic Statistical Business Process Model) that claims as the DDI Lifecycle (it was not the case for DDI codebook 15 years ago) to managing microdata from end to end, is that it includes from the start to the dissemination phase researchers. Thus documentation doesn’t remain the last thing to do after having produced microdata. Metadata have to be designed taking into account the reuse of microdata for scientific purposes which in turn increase the quality of metadata also for internal further use for the statisticians. In some cases, researchers can directly contribute to document microdata. This is particularly the case for administrative data. Since access to administrative data has been opened to researchers, many administrative data have been documented, sometimes with the help of researchers themselves (for example the tax data in France). Researchers’ input on the data quality itself seems currently less systematically integrated though there are numerous examples of how feedback from researchers can be used. One obstacle is the lack of organized process to involve the researchers and collect these feedbacks. Opening more access to European microdata as well as allowing easier transnational access to official microdata for the researchers would certainly allow them to provide more input on the quality of data and metadata which are crucial for comparability, a core issue for public policies.
|Title: <<< Quality in statistics and relations to the press and media >>>
As official statistical authorities, the statistical institutions have an important role in supporting a common fact-based perception of reality in society. Several of the traditional media also assume this role, and statistical institutions typically have formalised relations with the media. From a media perspective, however, the development in the past few years may be cause for serious concern. The internet has made everybody a publicist. Combined with the development in social media, it has changed journalism – and possibly also democracy. In the early days of the internet, the assumption was that it could contribute to strengthen our democracy and the public debate. Instead, the amount of knowledge has become immense, and it is often impossible to decode who is behind the information and what characterises the quality. The internet has become the most monitored place in society, and it is controlled by giant global players, such as Google and Facebook. In addition, these players have won an increasing share of the advertising revenue that was previously an important source of income for traditional media. The challenge in this is emphasised by the fact that many media have not yet come up with a business model that can finance serious journalism and also work in a situation where online news take up ever more space. A question is whether the national statistical offices have adjusted their relations with – and interaction with the media so as to make opportunities as well as pitfalls in the use of the official statistics known. Another question is whether the media’s use of the statistical institutions’ work on the quality and the dissemination of it is utilised to an extent that bears comparison with the costs. In the paper, Statistics Denmark presents various initiatives taken based on these challenges and the effects achieved.
|Title: <<< The dimensions of quality statistics in news. Strengths and weaknesses. >>>
Statistics are often associated with the notion of quality and credibility, as they are seen as a way of neutrally measuring and operationalising abstract and deductive reasoning. Simple statistical vocabulary such as averages, percentages, modes, medians or common values and proportions are included in the greater part of news reporting. In journalism, statistics are treated both as a credible source of information as well as hard facts. It is assumed that by inserting quantitative information in a story, reporters can contextualise a news item into the wider framework of societal trends while at the same time use the statistics themselves to strengthen the argumentation upon a certain event. All in all, the broad understanding is that by incorporating statistics journalists can make their stories better quality. However, as I ask in this presentation, does statistics really improve the ‘quality’ of the news stories? Does quality statistics translate into quality journalism? Quality journalism has been long debated among media scholars especially in recent years. Little has been said however on how statistics contribute to the debate around the notion of quality journalism. The empirical research attempted to “measure” the concept of quality and then contextualise it into the journalistic practice. The presentation is drawn upon empirical data from content analysis and in-depth interviews among journalists in the UK and tries to reflect on the strengths and weaknesses about the articulation of quality statistics in printed news. In addition, the research underlines the dynamics of the data-analysis routine and the numeracy of those journalists who use an alternative voice through mathematical tools.