Tuesday, October 31, 2017

A scary web article based on a couple of press releases

On October 27, 2017, at the Inc. web site, there was a scarily superficial article by Eric Mack titled Forget dying and public speaking: here’s the 47 things Americans fear more in 2017. First, the title is wrong - he ends by listing what more people fear and not what they fear more. In my October 29, 2017 blog post titled What do Americans fear most? Fear Scores from the 2017 Chapman Survey of American Fears I show what people fear more. Superficial journalism based on reading press releases is scary. There were press releases issued in 2014, 2015, 2016, and 2017. There also were blog posts in 2015, 2016, and 2017 all titled America’s Top Fears (year). Those fears were ranked based on the sum of percentages for Very Afraid and Afraid. Detailed results for 2014, 2015, 2016, and 2017 are in a set of .pdf files you can download.     

Eric’s first three paragraphs say:

“You've probably heard over the years that public speaking tops the list of things people fear most, freaking us out more than even the inescapable existential problem of death. But fear in America has grown and shifted dramatically over the last year or two, leaving death and public speaking buried beneath a long list of more pressing things to stress about. 

When Chapman University first conducted its ‘Survey on American Fears’ in 2014, ‘walking alone at night’ topped the list, followed by ‘becoming the victim of identity theft,’ ‘safety on the internet,’ ‘being the victim of a mass/random shooting’ and that familiar fear of public speaking rounded out the top five.

Earlier this month Chapman released the fourth annual edition of its fears survey and ‘walking alone at night’ has dropped all the way to be ranked number 56 on a list of 80 total fears that a sample of over 1,200 Americans were asked about. Public speaking ranks as fear number 52, while random mass shootings are at 35, safety on the internet didn't make this year's survey and fear of identity theft is ranked 14.”

When you actually look at the questions from the detailed results of that 2014 survey, you find a rather different picture than what Mr. Mack described in his second paragraph. Two questions were not on how afraid you are but how safe you feel, so they had answers on a scale running the opposite way from three others.

A question on page 18 under the heading for Safety asked:
“(Walking alone at night?) How safe do you feel:”
with possible answers of Not at All Safe, Somewhat Safe, Safe, Very Safe.

A question on page 53 under the heading for Fear of Criminal Victimization asked:
“How afraid are you of being victimized in the following ways? (Identity theft/Credit card fraud).”

Another question on page 18 under the heading for Safety asked:
“(On the Internet) How safe do you feel:”
again with possible answers of Not at All Safe, Somewhat Safe, Safe, Very Safe.

A question on page 55 under the heading for Fear of Criminal Victimization asked:
“(Being the victim of a random/mass shooting) How afraid are you of being victimized in the following ways?”

A question on page 66 under the heading for Phobias asked:
“[Public speaking] How afraid are you of the following?”

The press release for 2015 contained this warning about the 2014 rankings, which Eric apparently missed:

The researchers continue to improve the survey as its results and continuing interviews provide more information about fear, as well as how best to collect fear-based information. The second wave of the survey modified question wording such that all questions about fear use the same response categories: ‘Very afraid,’ ‘Afraid,’ ‘Slightly afraid,’ and ‘Not afraid.’ Consequently a comparison of fears between 2014 and 2015 should not be conducted without consultation with the researchers, who can explain the proper method for conducting comparisons across waves."

Presumably the proper method is to use a lot of hand waving and weasel words. Let’s take a look at the answers - detailed results for those questions across all four surveys. In the following comparison tables I’ve included both the sum for Very Afraid and Afraid, and the grand sum for Very Afraid, Afraid, and Slightly (or for 2014 Somewhat) Afraid. (Click on each table to see a larger, clearer view).

First, as shown above, for fear of Walking Alone at Night the percentages were (56.0), 16.0, 20,8, and 20,2. As soon as the question changed from safe to afraid the percentage plummeted.

Second, as shown above, 2014 was the only year with a combination for fear of Identity Theft/Credit Card Theft – which likely explains getting a higher percentage. For Identity Theft the percentages are [49.7], 38.4, 34.8, 41.7. For Credit Card Theft the percentages are [49.7], 35.7, 36.4, 42.5.

Third, as shown above, the 2014 question How Safe Do You Feel on the Internet (53.1) was replaced by Cyber-Terrorism 43.1, 30,0, 47.8.

Fourth, for Random/Mass Shooting the percentages are 24.7, 15.9, 26.4,30.8.

Fifth, as shown above, for Public Speaking the percentages are 25.3, 27.5, 25.5, 23.3. The mean is 25.4%, so the deviations from it are relatively small: -0.1% for 2014, 2.1% for 2015, 0.1% for 2016, and -2.1% for 2017.

Finally, for Corrupt Government Officials in 2014 the question was being Worried About (64.3%) rather than a fear by 56.1% in 2015, 58.9% in 2016, and 73.8% in 2017.    

The takeaway from this blog post is that to see what actually is going on, you need to look at the actual data not just press releases.

The image came from the collection of Images from the History of Medicine.

No comments: