Sensationalizing Science

Sunday, February 19, 2017

“Natural selection making 'education genes' rarer.” “Intelligence is being bred out of the gene pool.” Can you recognize which of these headlines is realistic or exaggerated? The answer is that each is a little of both — and they’re reflective of a disturbing trend. These and other current headlines are skewing real science.

In December of last year, Augustine Kong and colleagues of Decode Iceland released a large-scale data analysis of thousands of genomes purporting to show that “genes associated with level of educational attainment were decreasing”. The research itself made valid and well qualified conclusions, but as it spread, its message was diluted into hyperboles and ammunition for advocating preformed opinions.

The scientific study gained traction on news and social media sites under the above headlines. And that misleading online propagation reflects the dangers and pitfalls of scientific reporting in our era. As maximum clicks becomes the ultimate goal of so many websites, we should become hyper-aware of dubious science claims.

Vast differences in headlines covering the Icelandic paper illustrate what can go wrong when a scientific study is widely reported. The Guardian’s “Natural selection making education genes rarer,” is a much more measured claim than the Daily Mail’s “Intelligence is being bred out of the gene pool.” Such discrepancies happen when reporters refer only to secondary sources rather than the original paper. As this repeats itself across the web, all science is wrung out of conclusions and an oversimplification appears on search and social media feeds. Often, articles reporting on a scientific study never link back to their subject matter at all, but simply cite other similar websites.

These "end of the chain" headlines select the most striking results from the source article and compile them into something relatable or catchy. Suggestions that “intelligence” is decreasing is a far cry from “education genes” doing the same, but the former title attracts more people. Sensationalizing is found in many forms of online reporting, but because of its role in society, sensationalized science carries more serious consequences.

The headlines the Kong paper generated were so widely shared likely because they fed into a narrative that has captured the public for some time, despite minimal supporting evidence: that society is becoming progressively dumber. Judging from its resilience, people seem to be dedicated to this trope — perhaps because they never consider themselves to be a part of this trend.

In fact, a standard barometer for intelligence, IQ, has actually been shown to be increasing for decades, a phenomenon known as the Flynn effect. Environmental influences on something as complex as intelligence are profoundly important. The paper even concludes by stating “assuming that a similar magnitude of the Flynn effect is found in the Icelandic population, then it is clear that such environmentally induced increases of IQ scores more than compensate for, and indeed mask, any potential decline in the genetic propensity for IQ.”

Even when honest and reliable, journalism is still susceptible to exaggerated conclusions when the scientists may be fishing for a noteworthy result.

For example, the study claims it used “a significant portion of the population,” of Iceland and assesses a significant portion of the vast human genome. To the casual observer, it may seem inherently beneficial to have as large a sample size as possible, however, without getting into statistical theory, this is not true. Increasingly large samples make an insignificant p-value almost always significant, no matter the context. What is more meaningful in these situations is the size of the effect, which in this case was miniscule.

The threshold p-value for scientists and non-scientists alike has become somewhat of a gold standard for establishing credibility. However, in applications as massive as this research, the actual p-value approaches meaninglessness when interpreting the information’s significance.

The study addressed here isn’t wholly dishonest — it is actually quite cautious in its claims. But the point is, we need to verify our information when we stumble upon these widespread claims. At the very least, we need to read the abstract or claims of the original paper. That’s not always easy to find when reading online stories due to the endless cross-referencing of secondary and tertiary news outlets.

In the era of “fake news” one needs to foster awareness of “fake science” — or more specifically, sensationalized reporting on science— just as diligently. We may be even more obligated to do so for science because it holds such power in our culture, and its ramifications more serious. Regardless of who we are, we are all prone to blindly accepting claims that validate our beliefs, and so must strive to maintain healthy skepticism.