Hack pack on the rack

Journalists are prone to exaggerate the negative - no kidding. They also tend to neglect important details of research studies and cherry-pick the most sensational results - well, I never. Some may even work for news outlets with a political agenda and let that influence their choices of what to report.

Am I missing something here? Maybe the average American's sense of how the press works is radically different from this British reader's. And perhaps the US education system never mentions things that even junior school media studies seems to cover over here. But this book's main theses seem unstartling, to say the least.

They are, however, exhaustively documented, though through worked example rather than by any content analysis. Fair enough, this is a guide for the unwary public rather than an academic treatise - though the authors draw on academic literature in areas such as risk analysis. Most of the examples concern epidemiology and social science, and cover problems ranging from trends in Aids diagnosis and numbers of illegitimate births to crime figures and income inequalities. On the whole, the commentaries on stories are measured and convincing, although there are occasional hints of a rightwing tilt.

The authors take it for granted that only absolute, not relative, measures of poverty make any sense, for example. And they occasionally overstate their case, offering examples that do not demonstrate what is claimed. For instance, a story on the possible risks of drinking lots of coffee is said to show the alacrity with which journalists move from evidence of "possible links" to demonstrated causality. But the report cited does not make the move they criticise.

More seriously, lacking any quantitative data themselves, they drift into generalisations that belie the standards they upbraid others for failing to meet. When they tell us, for example, that "media exaggeration of risk ultimately stems from activist exaggeration of risk", they offer no real evidence for the assertion, and ignore most of the subtleties of risk interpretation and perception.

They also exhibit a naivety about the realities of media practice and the complexities of reader response that is embodied too in the guidelines for media science reporting recently promulgated in the United Kingdom by a collection of well-meaning but misguided scientific bodies, including the Royal Society and the Royal Institution. They counsel against "disproportionate" coverage of preliminary research, and tend to assume that readers only ever encounter one story about any given topic, and believe what they read.

The combination of a neglect of serious analysis of reader behaviour and the absence of any basis for judging the prevalence of the faults the authors find in media performance leaves one with mixed messages. They concede early on, for instance, that "in some sense everyone knows" that news is highly selective and can never be a complete or objective account of the world. On the other hand, they use a discussion of women's lifetime incidence of breast cancer to illustrate that "the print media have often excelled at explaining the realities of risk".

The result is a book that helps its readers - the authors call them "news consumers" - view the details of media reports of research sceptically, but fares less well when the sceptical eye is turned on its own advice. And it left me feeling frustrated by the presentation of a large collection of examples that add up to less than their sum. The research resources that went into this book might have been better spent on a more detailed account of a smaller number of cases. These could have been used to address some more penetrating questions about the dynamics of news reporting of controversial issues, for example, that the stack of snapshots assembled here cannot. A missed opportunity.

Jon Turney teaches science communication at University College London.