On the importance of data literacy…

Published Categorized as Data Literacy, Musings No Comments on On the importance of data literacy…

Well, checking your sources at least..

I once presented on Psuedo-Profound Bullshit (yes, it’s an academic term) and people’s receptivity to bullshit. To my own, and my teams amusement, I took them for a ride and left them scratching their heads as to whether I had indeed just fed them bullshit about psuedo-profound bullshit.

Months later, scrolling LinkedIn I came across an article claiming scientific research into the effectiveness of magnetic therapy in curing chronic pain. My bullshit detector alarm sounded.

As a person living with chronic pain, the spread of misinformation and disinformation on chronic pain relief is of personal importance to me. So, I clicked to investigate.

The article linked to research papers to support its claim. That sounds somewhat credible, but knowing how research design and methods are often flawed, I followed these links to read more. The first link claiming a static magnet could relieve pain in 95% of cases led me to an abstract. Nowhere here could I read that the study found 95% alleviation, it did however state it used a 95% confidence interval in its statistical analyses… so, perhaps it was in the paper itself. I googled but was repeatedly paywalled until I found this link. Nowhere within the article did it say that there was 95% alleviation of pain, but it did claim 53% which was of statistical significance.

So it appears that the author of the article intentionally or unintentionally reported a confidence interval as a research finding in support of their sales agenda.

Hmmm…

I see this so often, not in sales pitches but in articles by reputable business journals, where an author justifies their opinions with previous “studies”, but when following the trail of studies you are left with circular arguments, unreplicated findings, and inferences. Nothing sound that should be allowed to sway the minds of a wide-reaching audience that will not be fact-checking nor using critical thought.

It’s not a new phenomenon. But hey, in the context of data viz, it’s important to understand your subject matter to avoid a misrepresentation of truth.

Leave a comment

Your email address will not be published.