Opinions on science derailed by poor online sample?

Scientific American and Nature recently joined forces to poll readers around the world about their opinions of science. The findings include opinions about science and politics, climate denial, nuclear power, the flu and more.

While this data seems interesting, it might be questionable due to the sample:

More than 21,000 people responded via the Web sites of Nature and of Scientific American and its international editions. As expected, it was a supportive and science-literate crowd—19 percent identified themselves as Ph.Ds. But attitudes differed widely depending on particular issues—climate, evolution, technology—and on whether respondents live in the U.S., Europe or Asia.

So the findings may really be about the opinions of a more “supportive and science-literate crowd” rather than a true representation of international opinion. This is a common issue with open online surveys: it is very difficult to get a sample that is representative of a larger population.

From controversial opinion piece to full length book

I was unaware that this was a common phenomenon: write a controversial op-ed in a major newspaper, receive a book deal, and then produce a book that is much too long and that doesn’t argue much. David Bell describes this process:

The syndrome has become all too common. A provocative op-ed piece appears in a major newspaper (for preference, The New York Times). Its logic is fragile and its evidence is thin, but the writing is crisp and the examples are pungent, and the assault on sacred cows arouses a storm of discussion (much of it sharply critical, but no matter). It goes viral. And almost immediately, publishers comes calling. “This should be a book,” they coo, and the author, entranced by a bit of sudden fame (not to mention, perhaps, a decent advance), eagerly agrees. He or she sets to work, and soon enough the original 800 words expand to 50,000. But far from reinforcing the original logic and evidence, the new accretions of text only strain them further, while smothering the original provocations under thick layers of padded anecdote, pop sociology and oracular pronouncement. Call the syndrome Friedmanitis, after a prominent early victim, the New York Times columnist Tom Friedman.

I wonder if I have been a victim of this process. I have read a number of non-fiction books where I thought the argument was thin and the argument could have been effectively made in just a few pages. One problem may be a lack of data – opinion books are difficult to sustain as they often jump from one opinion to another without providing sufficient evidence for the claims being made.

From the book publishers perspective, this process makes some sense. Perhaps the hope is that the op-ed author has more to offer; that if given more space, they can develop a much more substantive argument. Since it is difficult to predict which books will succeed once published, an op-ed that generates attention may look like a sure thing.

At the same time, these op-eds can quickly invoke many criticisms within hours of being published online. By the time a book is released that is built around the same topic, it may be too late to make the argument again (particularly if it is badly argued in the book).