Our central question, so far, has asked about the nature of the engagement between journalists and statistical data in the pursuit of quality. In so doing, we would want also to know if the quality statistics automatically lead to quality journalism. If so, we ought to know if the information quality then translates into quality journalism. Does the nature of a statistic's source affect the news reporting? What is the purpose of statistics in news reporting? Do journalists emphasize a certain type of statistics? What statistics sources do journalists use most often? And, how does the audience engage with statistically driven stories?

To answer these and other questions, we used a mixed-method approach, in which we have triangulated qualitative and quantitative methods as this ‘is major research approach’ (Johnson, Onwuegbuzie & Turner, 2007). The importance of such a triangulation is the validity of the results that can lead to a more balanced and detailed answer to the research questions by also comparing and contrasting different accounts of the same situation (Turner & Turner, 2009). The aim was to develop a ‘practical theory’ that would help to rationalize the issue under scrutiny (Altrichter, 2010; Altrichter, Posch & Somekh, 1993). We included content analysis, semi-structured interview, close reading and focus groups, which allowed us to carry out a multilevel assessment of the data. The overall mixed method used in this research should be understood in a broader ‘cross-sectional design’ (F. L. Cook et al., 1983; Johnston & Brady, 2002), which allows a combination of quantitative and qualitative research.

Therefore, this chapter presents the analysis of the data collected followed by a discussion of the research findings that resulted from each method. It provides a detailed account of the findings, in the hope that these results will elucidate the uses of statistics in articulating the five quality dimensions in news reporting.

Overall, our findings suggest that journalists tend to use statistical information as a tool to fulfil their deontological expectations of producing quality journalism. However, as it became clear from the interviews, one of the underlying motivations seems to also be the need to achieve credibility and authority, which entails a certain degree of building up the ability to persuade by means of trust.

Over the past years, governments and corporations have come to develop media strategies to ‘manage’ the dissemination of statistics within the public. As we have discussed in this book, this was part of a long history in using numbers to assert social control and that in the twentieth century took the form of cybernetics. In more recent times, this has meant enshrining statistics and data into daily life by means of the information and communication technologies to be used by individuals, couples, families and communities. In so doing, they have dedicated immense resources towards ‘controlling’ narratives and interpretations of statistics released in the public sphere.

Moreover, mediatization practices to shape the way these numbers go out into the public are now at the centre of the controversies and issues that journalism confronts. This ‘mediatization of statistics communication’ (Lugo-Ocando & Lawson, 2017) is one which is understood to be not a policy-making process directed by the media but overall a policy-process in which publicityseeking activities and political decision-making become closely interlinked (Cater, 1965; T. Cook, 2005).

Statistics are far from being a neutral object in society and have their own politics and ideologies. They are also fundamental signifiers in the creation of social reality (Dorling & Simpson, 1999), displaying a politics of their own. Not that we ascribe agency to the output of a mathematical equation but that we see the equation itself as a human creation that has a history, a meaning and an intent. Over the years, as we have seen in this book, these numbers have gained a power of their own. From defining budgetary priorities to determining who can receive aid or buy a house, statistics and data, in more general terms, dominate human existence in many ways.

However, with the increasing presence of Big Data and controlling power of algorithms, we are entering into new uncharted territory, No longer is it just about the state or corporations controlling the production and analysis of statistics, but also about the active management of these numbers by means of mediated representation in the public domain. Recognized today as a process

The assumptions that quality can be only asserted through numbers hold both within journalism and government. Particularly, as these numbers bring a sense of impartial and objective assessment to both formulate and evaluate policies. Moreover, at the centre of this assumption is the deep-rooted belief that statistics can bring about the type of scientific-based knowledge, transparency and trust needed to implement and analyse government policy. Nowhere was this more evident than during the Tony Blair-led New Labour government in the United Kingdom (1997–2001), when evidence-based policy became central to the way politicians sold their own agendas to the public (Hope, 2004, 2005). To be sure, the Command Paper released under the title Statistics: A Matter of Trust (1998) made it clear that

In this sense, the UK Statistics Authority has adopted a structure broadly similar to that of the European Code, which sets out a number of high-level principles, each of which is further amplified by a series of more detailed practices (or ‘indicators’ in the European Code). Also, the UK Code of Practice for Official Statistics and the assessment programme that follows have been informed by, and are consistent with, both the UN Fundamental Principles of Official Statistics3 and the European Statistics Code of Practice.

According to Mark Pont, member of the board of directors of the UK Statistics Authority, the European Code has proved an effective basis for the international process of ‘peer review’ and ‘the Statistics Authority believes that a similar approach will provide a sound foundation for the Statistics Authority's quality assessment function’ (Code of Practice: 8). One of the strong points of the Code is its emphasis on the role of the user, and the need for statistical producers to consider the wider use that is – or may be – made of statistics. In addition to meeting specific policy needs within government, there is increasing demand by people working in research, academia and journalism for statistics in many aspects of social and economic life.

There seems to be uniqueness to the logic of journalists’ way of thinking. That is, news reporters have their own way of interpreting the world outside and developing a close system upon which social reality is constructed in the public sphere. The power of journalism as a political institution to define the terms of reference for discursive regimes and narratives that frame people, places and events has been analysed and dissected by sociologists, media scholars and political scientists. However, insufficient engagement has happened with logic and cognitive psychology, which we believe is essential in the analysis regarding the use of statistics in the context of journalism.

To be sure, early calls by Holly Stocking and Paget Gross (1989) highlighted the failure of journalism studies to employ research by cognitive psychologists with regard to the kinds of errors and biases that can negatively alter the processing of information. According to Stocking and Gross, there is a lack of academic research that looks explicitly at the way journalists mentally process information when reporting and gathering the news because there are a number of specific errors and biases that ‘have been found in a variety of professions across a variety of tasks’ as well (Josephi, 2009).

Interestingly, if errors and biases are examined from the perspective of cognitive psychology, several of the errors and biases are also studied in the branch of philosophy known as Logic. Philosophy of Information also belongs to this branch. As Elliot Cohen writes:

In information processing, factors like prejudices, prior expectations, values, poor insight, visual conditions and emotional stress affect the ‘quality’ of this inferential reasoning. Stocking and Gross are clear in warning journalists about committing ‘the eyewitness fallacy’, that is, the fallacy of overestimating the reliability of eyewitness reports as compared with other sources of information.

To understand how journalists use statistics to achieve quality, we need first to contextualize news reporting practice within the wider ideological context of professionalization. In this sense, Professor Mark Deuze (2005) has argued that the professional identity of journalists is held together by an occupational ideology of journalism. Historically speaking, a person's occupation exerts important influence in determining their role and their family's position in society, and in the past even the place the individual would live (Mack, 1957), which also underpins the more general ideologies that these individuals embraced as part of their beliefs (Dibble, 1962). However, the great reconfiguration of society that took place in the West due to deindustrialization, the growth of the service and financial sectors and the creation and incorporation of a set of Information and Communication Technologies (ICTs) that alter most people's daily lives have meant that all professionals now operate under very different occupational ideologies.

These transformations can be encapsulated within the notion of the Information and Network Society about which several authors have referred to over the years (D. Bell, 1973; Castells, 2011; Drucker, 2012; Mattelart, 2000, 2003). It is a concept that reflects specific trends in the capitalist society and that has had an important impact in particular areas such as the media industry. Perhaps no other profession has had to endure such significant changes as journalism given not only the dissemination of the political economy that used to sustain the media industry but also the development of a completely new media ecology that has transformed working conditions and professional practices around news reporting.

Therefore, the use of statistics in journalism should be understood within the context of the Information Society. The notion of the Information Society took shape during World War II with the invention of ‘thinking’ machines (Dyson, 2012). However, it only became a standard reference in academic, political and economic circles from the 1960s onwards, thanks to promotion of the idea by scholars such as Daniel Bell (1973, 1976), who is recognized as the foremost writer on the Information Society, developing a robust argumentation around the subject from the 1960s to the 1990s (Duff, 1998).