About the SA Blog Network

The SA Incubator

The SA Incubator

The next generation of science writers and journalists.
The SA Incubator HomeAboutContact

For Up-And-Coming Science Journalists, Understanding Statistics Has Never Been More Important

The views expressed are those of the author and are not necessarily those of Scientific American.

Email   PrintPrint

Maths on a blackboard

By stuartpilbrow (from flickr).

“When Bill Gates walks into a bar… the average salary goes up.” – Popular geeky stats joke.

I once heard a science editor at a rather well-known publication say, in public no less, that she has no idea what p-value* means. This came as a shock to me, a then-relative newcomer to the science communication sphere. Why shock? Because, as I once wrote, statistics “carries the purity of the sciences on its shoulders.” Indeed, as the scientific method is based entirely on statistics, not having a decent grasp of it leaves science reporting prone to serious mishaps.

Of course, I realised fairly quickly that not all science communicators are inept when it comes to statistics. But the countless examples of statistics misinterpretation by the media when reporting scientific findings and discoveries** (looking at you Daily Mail) clearly shows that there is much room for improvement.

An upcoming book, written by members of the Scilance community, may help out in this regard. An excerpt from The Science Writers’ Handbook: Everything You Need to Know to Pitch, Publish, and Prosper in the Digital Age, published today exclusively by The Open Notebook, provides science writers with an important primer on statistics.

The excerpt is divided into three sections: “The Uncertainties of Uncertainty,” “Seeing the Story in the Stats” and “A Science Writer’s Statistical Phrasebook.” All three sections are important although I would like to highlight the second and third sections here. These two sections specifically draw attention to and attempt to explain potentially tricky statistical terminologies or concepts like percentage points, confidence intervals, absolute and relative risks and the infamous p-value. Admittedly, it’s all very basic and the explanations are sometimes a bit confusing but the sections do provide a decent base for science communicators to build on.

Up-and-coming science writers should be particularly receptive to statistics in the light of the “big data” era we’re entering. Scientific papers full of data and complex statistical methodologies will become increasingly prevalent and relevant in the near future. Consequently, adequate and critical science journalism will require a good statistics grounding more than ever before.

* p-value, as the excerpt enticingly explains, is the “likelihood that the observed test result happened by chance. A low p-value means the results were significant and unlikely to have occurred by chance.”

** Science is not the only beat suffering from statistics misinterpretation. Sports, for instance, is currently going through a data revolution leading to a staggering amount of erroneous statistical analyses by sports writers who are casually interchanging correlation with causation.

Khalil A. Cassimally About the Author: Khalil A. Cassimally is the Community Coordinator of The Conversation UK. He's also a science blogger. He hails from a tropical island and is a happy geek. Subscribe to his updates on Facebook and Google+. Follow on Twitter @notscientific.

The views expressed are those of the author and are not necessarily those of Scientific American.


Rights & Permissions

Comments 3 Comments

Add Comment
  1. 1. jdottan 12:00 pm 04/30/2013

    I would actually preferably define the p-value as “the probability of the test result happening assuming that the null hypothesis is true.”

    Link to this
  2. 2. kausikdatta 12:07 pm 04/30/2013

    Great piece, Khalil. As a researcher in a biomedical discipline, I cannot emphasize enough the importance of understanding the basics of statistics for scientists, as well as anyone who communicates scientific data to others (such science writers, educators and journalists). Too many people still treat statistics as a magical box, believing that their random or biologically implausible observations can somehow be transformed to significant facts by statistical legerdemain. Statistics involves mathematical operations with numbers; what makes it important is the interpretation of the computed value, and this cannot be done properly without a training in the basics of statistics. It is important to remember that the use of inappropriate statistical operations or incorrect interpretations only serves to reduce the value of the observations made.

    Link to this
  3. 3. plswinford 3:09 pm 04/30/2013

    I am in favor of statistical education since I keep seeing claims like “this process doubles the chance of you getting cancer”, without the useful information that the chance went from 1 percent to 2 percent, but the cost per patient went down 50,000 dollars. Thus ignorance continues its high batting average.

    Link to this

Add a Comment
You must sign in or register as a member to submit a comment.

More from Scientific American

Email this Article