23/6/2016

News values in data journalism, or: zOMG you won’t believe what data is doing to our news…

 

Data is flowing into (and out of) our newsrooms like never before – I trust this claim will seem fairly uncontroversial to readers of this blog. The numbers come in many forms; from audience metrics, to spreadsheets and databases hoovered up during news-gathering, to the converged formats we interact with on our smart phones.

But what are the consequences of these data flows on contemporary journalists and journalism? Judging by the latest coverage, it's not good news.

Recently we have learned that 'contributors' (columnists, once upon a time) at City AM will be paid by the pageview, rather than by the quality of their writing. Earlier in the year, management at Trinity Mirror were forced (by fear of industrial action) to abandon setting their journalists individual online growth targets. Even the alternative press are seemingly open to payment by the click.

Some may say this is merely the logical conclusion of a networked news culture.  Back in 2009, a new expression was coined to describe this phenomenon: 'the culture of the clickstream'.  In What's Happening To Our News, Andrew Currah offered anecdotal evidence that in the modern networked newsroom there is a "growing tension between editorial values and knowledge of what will actually generate revenue".

The danger, it is argued, is simple: chasing numbers (in this case pageviews), must necessarily drag journalism down to what in Louis Brandeis' terms is a classic 'race to the bottom'. Fluffy bunnies, listicles and psychologically manipulative headlines will inevitably crowd out 'proper journalism', so the reasoning goes.

So, are the data Jeremiahs right? 

To understand what is really going on in our news, it is necessary to take a longer view than a single (largely anecdotal, and arguably technologically determinist) study. If we wish to study this issue seriously, we must engage with an area of academic research whose decades-old pedigree can tell us much about the formulation and structure of our news: news values.

What are news values? They are the mythical subjective, largely unwritten criteria that guide journalists in deciding which of the countless potential stories they should cover, and which to prioritise in their coverage.

For journalists, these things seem to be second nature; but academics have long observed that news values are more to do with the adoption of institutionalised values and systems than they are the manifestation of instinct or a 'nose for news’.  

News values

The earliest studies of news culture were conducted during an era when researchers in the social sciences took a hubristic approach to the objects of their study; assuming, for example, that culture can be 'captured', employing 'scientific' methods that seek to drive out, rather than engage with difference and nuance, and often establishing an unequal relationship between researcher and subject. 

The most celebrated of these early studies was carried out by Galtung and Ruge whose theorisation of news values comprises a list of 12 factors,  which were arrived at by means of analysing content from four Norwegian newspapers.

Over the years, limitations to and shortcomings in this approach have been raised:

  • The approach taken says nothing of the deliberative process that goes into formulating news values (and so it doesn't say much about the concept as we understand it today)
  • The approach may tell us more about story treatment than it does about why stories are selected in the first place
  • This study says nothing about news values in non-print journalism
  • As with all normative approaches, these values are vulnerable to obsolescence over time
  • Audience effect was not considered

In seeking to address some of these issues, a less structured, more consensual school of inquiry into news values emerged during the 1970s and 1980s, in the form of organisational studies of the newsroom. This approach sought knowledge about news values from the 'thick description' of field studies of news culture, as practised, experienced, and lived in the newsroom.  These studies identified a guiding tension in news values; the journalist is caught between professional and organisational values, each of which exert pressure on the actual processes of gathering, negotiating and publishing news, and which fundamentally shape our news.

This 'cultural turn' begat further studies of news values, giving rise to an ever-splintering range of competing voices and approaches to understanding the phenomenon.

As many developing nations established their independence during the second half of the twentieth century, the western approach to reporting, imported in earlier times, was increasingly found to be wanting.  This in turn led to the emergence of the 'development journalist'; whose news values were framed by ideals such as consensus, tolerance, partnership-building and social improvement (as opposed to reductivism, over-reliance on news personalities, and the simplistic treatment of complex issues that often crowd out western news).

More recent research suggests that female journalists differ from their male counterparts in terms of the values they express in their reporting; women tend to be more concerned with context and experience than end results, and often find their approach to be at odds with dominant (male) news values in the newsroom.

So if cultural issues may be seen to influence news values, then what is the effect of the contemporary culture of data journalism on our news? And how can we investigate it?

News values in data journalism

Today the 'culture of the clickstream' finds expression in various forms; in the rise of automation and algorithms in our news, in increasing ‘churnalism’, and perhaps most controversially (for now), in the rise of clickbait.

But the problem with addressing these phenomena often lies with the terminology itself – clickbait is a rather fuzzy concept.

The standard definition of clickbait is: “content whose main purpose is to attract attention and encourage visitors to click on a link to a particular web page”. So far so good – but this tells us little about the negative aspects of the phenomenon.

Alternatively, the concept may be defined as: “online material (such as headlines) designed to make readers want to click on hyperlinks especially when the links lead to content of dubious value or interest”. But this definition is clearly premised upon subjective interpretation; what is of ‘dubious value or interest’ is, frankly, anyone's guess.

In a departure from the approaches taken above, clickbait may alternatively be conceived of as: “...also a primary technique used to make phishing communications attractive to the unwary recipient".

These three definitions span a spectrum of ethical behaviour; from the common-sense approach in which journalists seek to maximise the yield of their online labours, right through to behaviour that may be denounced as criminal, let alone unethical.

But beyond definitions, there are further problems inherent to studying clickbait as a manifestation of news too - its visibility needn't necessary tell us much about the 'culture of the clickstream' in so far as it tells us little about the process of selection and negotiation that is central to determining what our news is, and how it is prioritised.

Those who fail to learn from history…

Some of the arguments that arise in debates about ‘the culture of the clickstream’ merely represent re-visitations of discussions raised long before the rise of online news.

Tabloidization is today recognised as a global phenomenon – it is: “the transformation of news, literature, etc., into a popularized, lurid, and sensational form”. Concerns about this phenomenon in news first found voice in the late 1990s, when the concept was defined to encompass processes relating to both the subject matter and treatment of news.

Though some reject the term on the grounds that it is sociologically unsound, and so better suited to comment in newspapers than serious academic study, nonetheless two competing approaches to the phenomenon arose.  On one hand, it is argued that tabloidization represents a 'dumbing down' of journalism's public function; a squandering of its fourth estate legacy.   On the other hand, it may be countered that such hostility to the form merely represents snobbery based on hostility to popular culture.

So does ‘the culture of the clickstream’ represents a contemporary manifestation of tabloidization? Perhaps this question is best answered with reference to an even wider historical context, going back through various iterations of (essentially) the same debate about standards in news we've been having since the emergence of the 'new journalism' in the late 19th century. There are, it is argued, four phases in this process – all predicated upon binary comparisons of 'good' and 'bad' journalism:

  • 1880s – 1920s: Sensational 'yellow' journalism versus the old-style 'Top Hat' journalism
  • 1920s – 1950s: Mass circulation news versus the elite press
  • 1960s – 1980s: Unscrupulous tabloids versus virtuous tabloids
  • 1980s – 2000s: Mass-market (image-based) television versus (text-based) broadsheets

It may be helpful to think of the 'culture of the clickstream' as a fifth phase in this recurring theme. By thinking about the issue in these terms, it helps to broaden out the debate about data in our news, requiring that we acknowledge (and consider) a counter-veiling form of online news culture.  In decrying clickbait (and 'the culture of the clickstream’), it is possible to lose sight of the value in using data to inform newsroom functions.

In my study of Search Engine Optimisation (SEO) in newsrooms, I found that SEO practice is circumscribed by existing professional norms and organisational routines. SEO was only found to influence editorial decision-making directly in one of the four news organisations I studied; and even then, this was of little value in terms of day-to-day traffic. It is feasible that ‘hard to find’ stories may become non-privileged when SEO becomes the prevailing influence on the way online news is sourced, produced and disseminated.  The pressure that SEO may exert over news style, and news values, (on behalf not of readers, or news organisations, but according to the algorithms of third party news aggregators) requires that we check regularly. Nonetheless, these findings represent a challenge to the notion that the 'culture of the clickstream' represents an existential threat to news values today.

But the 'culture of the clickstream' is a reaction to the influx of data into modern newsrooms in a single, rather narrow context too; it says little about the innovation of data journalism in today's news.

In a newsroom study of interactive graphics, I found that some non-journalistic professional values represent a challenge to decision-making in the networked newsroom, and that non-journalists (including graphic designers and programmers) openly question the coverage and treatment of certain stories.  Interactives are infinitely adaptable in news and features content; their use confounds those classificatory distinctions established between news 'types' found in communications studies literature.

In a recent interpretation of these findings, it has been proposed that the audio-visual nature of interactive news represents an emerging news value. But I believe it is possible to take this argument further still.  Data journalists sometimes opt for graphical forms of lesser numerical integrity than 'best practice' might otherwise permit; in-house research often suggests that by doing so media stand to alienate less of their audience, and so may communicate the story more effectively (to a wider audience).  This surely suggests that there is potential here to engage the public in civic and political matters in ways that may benefit society more generally, something that requires further study. 

Might data become a news value in the modern newsroom?  Studies have shown that the news values of TV journalists are driven by the availability of news footage – is it possible that future journalists' prioritising of news will be determined by the availability of data. If so, what might this mean for areas of life that are under-represented by data, that are un-audited; that are analogue?

Picture Superiority Effect, which emerged from empirical studies in cognitive psychology, proves that concepts learned using pictures and text are recalled more easily and more frequently than their equivalent expressed in word-form only. The explosion of interactive graphics in online news may therefore represent an opportunity in civic and political communications. When the public are able to remember the numbers in the news for longer, news becomes less and less merely tomorrows fish and chip wrappings – might this have ramifications for how well informed are our citizenry, and the political realm more broadly, in future?

It is clear then, that there are far more questions about the culture of data journalism in our newsrooms, than there are answers; and many of them have little direct bearing on 'the culture of the clickstream'.

Conclusion

Is data making our news better or worse?

The simple answer is we don't know.

At a conference at Cardiff University last year, keynote speaker Professor Stephen D. Reese (of the University of Texas at Austin) called for further ethnographies of the networked newsroom; to help us make better sense of the effects of data on our news.

This is all well and good, but this call comes as researchers are finding it increasingly difficult to gain access to newsrooms. My own experience is that access can be partial; some questions and lines of inquiry (especially those to do with commercial matters) were blocked outright, and each interaction had to be carefully negotiated (sometime by offering to share information with journalists as collaborators and partners, and sometimes by offering skills and advice in my own technical specialism).  In the end, the information I missed out on was secondary to understanding decision-making within the newsroom, nevertheless it speaks to a wider issue – the need for a new settlement between media and academia, based on mutual respect. 

We need more collaboration between academics and newsrooms.  As academics we need to justify and make good on that access, rather than treating journalists as if they are an ideological problem, or mere matter in a petri dish (as has happened all too often in the past).  We need more participant observations, involving academics with useful transferable skills, so that news media can benefit from knowledge in the academy, just as academics benefit from these studies.  And we need more practitioner academics whose research is based on the principle of collaboration, mutual help and respect.

It is a bit too early to say what the effects of data on our news are. This may seem like an unsatisfying answer, but if we can establish a new settlement between media and academia, then future studies will hopefully take us closer to the truth.

Image: Dimitris Kalogeropoylos

Comments