fake news

The issue that got under my skin today is how sensationalism sells and how this has misled us and distorted “news.” As more of us rely on screens and social media to alert us to important world issues, yellow journalists are setting up fake virtual offices to create inflammatory, sharable articles. Each click-through brings money into their pockets from ads. That means a gullible public pays cons to misinform us. The cons recognize the more outrageous the claim, the more attention they get, and the more money they make. They have learned that repeating inflammatory statements over and over can lead to their adoption as facts. Does an intentionally groomed and misled public threaten American democracy?

The First Amendment

Our founding fathers created our governing system with the expressed intent of rule by the people. To protect the freedom and power of the people to be informed and vote, they wrote the first constitutional amendment protecting the free exercise of religion, the freedom of speech, the freedom of the press, the right to peaceably assemble, and the right to petition for a governmental redress of grievances.

They believed that the free flow of information to the American public from a press that serves as a credible watchdog keeps powerful officials in check. This was considered critical to American freedom.

Sensational Headlines

I started GetKidsInternetSafe (GKIS) because, despite being somewhat tech-savvy with great personal and professional resources, I was struggling to create a workable plan for managing my children’s technology use. I’d take a stab at this and that but either get distracted along the way because the plan was too cumbersome or become completely frustrated and throw up my hands. Ultimately I’d end up doing very little and feeling chronically guilty.

My anxious search was exasperating partly because the Internet is flooded with sensational headlines that exaggerate risk! Even responsible reporters sometimes bypass credible experts in favor of interviewing somebody willing to be provocative rather than factually responsible.

Then another article would counter-react with the opposing but also exaggerated viewpoint. Furthermore, sometimes even credible experts are inaccurately quoted. I was so frustrated trying to get sensible information that had persuasive evidence backing it, that I’d just give up in defeat.

As I turned to my family, friends, and colleagues, they just looked blankly back and said, “Yep! Exactly! I’m paralyzed too. Tracy, if you can’t do it with your expertise, then we’re doomed.”

Doomed? No way, not when it comes to kids. So from there, I sought out the smartest, most creative, most energetic hotshots I knew and launched GKIS. I decided to do the hard work and comb through the psychology research, news articles, academic theories, and tried-and-true parenting strategies to create a comprehensive, easy parenting course that works.

Religion and the Internet

As I was researching for my book Screen Time and Mean Time and my GKIS parenting courses, I ran across an article titled, “America’s Less Religious: Study Puts Some Blame on the Internet,” by Elise Hu on NPR. The article states that as Internet use has grown, people have become less religious. A study by computer scientist Allen Downey reportedly “found a causal relationship among three factors – a drop in religious upbringing, an increase in college-level education, and the increase in Internet use.” A causal relationship? They were saying that less religion causes more internet use and more education. That makes no sense. I’ll tell you why that’s a false claim.

The Experiment

I teach my university students that the only type of study that yields evidence about a causal relationship is an experiment. An experiment is conducted in the laboratory where you take two groups of subjects and expose one to the experimental variable (the independent variable)  and then measure the outcome for both groups (that measurement is the dependent variable). If the groups’ outcomes are the same, your independent variable did not have an impact. If they are more different than could happen by chance, then your independent variable did have an impact. You have to do other things to make sure your experiment is high quality, but the main point is that you have to act upon your subjects – not just survey or observe them.

Observational studies only suggest that there may be a relationship, but they cannot rule out unidentified outside (extraneous) variables that may contribute to the relationship between the study variables. More to the point, because they’re observational, these studies have no way to determine which – if any – of the variables they’re studying is causing the outcome and which is being affected. Causal claims are impossible to support if you have not conducted an experiment.

Yet this NPR article suggests that, since Downey can’t think of that third variable, there must not be one. And with a logistical regression method, voila! He can claim Internet use has killed religion! Pffft. This is a blatant misrepresentation of the data. Yes, based on the statistics those three variables appear to be related. But we don’t know if increased Internet use led to less religion or if less religion led to increased Internet use. A fourth unrecognized variable may have led to both, like maybe Internet use led to more screen time which led to more Facebook which led to more overwhelm which led to less leaving the house which led to less religiosity. I mean honestly, we don’t know until we integrate all variables or, better yet, design an experiment!

In this case, it is unclear to me if the researcher made these irresponsible claims or if the reporter misreported, but the article is flat wrong in its conclusion and misleading to the readers.

Dr. Rochelle Tractenberg

As a social scientist, I was irritated. So, in my ranting mood, I sought out my friend brilliant Georgetown double Ph.D. biostatistician Dr. Rochelle Tractenberg, and asked her to give us a quick and informed comment about her reaction to the NPR story about this study. You can’t get a more credentialed expert with statistics than Dr. Tractenberg. She responded:

My friend and colleague Tracy Bennett asked me to contribute some comments in response to the frustration she felt – and felt she needed to share with you! – after reading the media coverage of what seems, based on this coverage, to be a very weak study. I am a scientist myself, and I coach other scientists in the design and write-up of their work. I completely agree with Tracy about this media coverage!

As a person who sees “scientific articles” every day of the week, I am frustrated by how any “science” gets talked about in newspapers, on TV, blogs, “the Internet,” and radio. Probably the most common problem is that people outside of the field of science often believe that, if a scientific article made it through to publication, then it must be good, true, or right. That is just not true! Just because a paper was published in a “peer-reviewed” journal doesn’t make it good, correct, or meaningful.

I think a wider problem is that

_____________________________________________

journalists should not be trusted to sift through what was published to bring “what matters” to the public’s attention.

_____________________________________________

Just because a news outlet covers a paper’s having been published doesn’t mean the paper is accurate or even interesting. You might not know that new scientific papers are published every second of every day; probably TWO others were published in the same week on the exact same topic, with the opposite <maybe even the same!> results as whatever paper the media are covering that day. The question the reader must ask is, WHY WAS *THIS PAPER* CHOSEN FOR THIS REPORT? Very probably it was chosen because “it sounded interesting!” The CHOICE to cover a particular paper reflects the journalist’s interest and not the importance of the paper – not for its field and not for the public.

The purpose of scientists in publishing their work is usually to test theories and contribute new knowledge, but journalists’ purposes are to attract readers (if you have ever read a scientific paper, you can see plainly that it was NOT written to attract readers!).

_____________________________________________

That means that people who get their “science” from journalists do not actually get “science.”

_____________________________________________

Science moves in truly small steps and very few newsworthy “breakthroughs” ever occur, although news media make non-scientists believe that they happen often. Exceptions may come from fields like archeology or astronomy – where observations and not experiments are reported; if a new dinosaur was discovered and reported in the general media then it probably IS important!

Scientific papers follow very specific rules – including that they must consider the work that was done before and how their new results fit with those older results (whether the new results agree or disagree with the dominant theories). They must also always describe the limitations of their work and suggest what more might need to be done in future research.

Media coverage follows very different rules; the article Tracy found was probably chosen because it is provocative – it could not have been chosen for the NPR story because it is true (impossible to tell!) or well done (it wasn’t). I would go so far as to say that all that a reader really can “learn” from a news story on something scientific is that “that research is going on.” A reader should NEVER infer from a news story about research that something important for daily life has been learned.

Rochelle E. Tractenberg, Ph. D., M.P.H., Ph.D., PStat®
Director, Collaborative for Research on Outcomes and -Metrics
http://crom.gumc.georgetown.edu
Associate Professor
Neurology, Biostatistics, Bioinformatics & Biomathematics, and Psychiatry
Georgetown University Medical Center

[Read more…] about Is it a Scientific Finding or a Sensational Headline?