Friday, October 30, 2015

According to the 2015 Chapman Survey of American Fears, adults are less than Afraid of federal government Corruption and only Slightly Afraid of Public Speaking

In the 2015 Chapman Survey on American Fears 1,541 adults were asked around ninety questions that all began with:

“How afraid are you of the following?”

They were asked to choose from one of four answers (fear levels):

1]  Not Afraid

2]  Slightly Afraid

3]  Afraid

4]  Very Afraid

In the Chapman blog post about the survey on October 13th titled America’s Top Fears 2015 there was a Complete List of Fears ranked by the sum of the percentages for Afraid and Very Afraid (and also shown alphabetically). Although the text says there are 88 fears, there actually are 89, since Dying is listed twice but actually refers to two slightly different questions, the second being Dying Alone.


That blog post also discussed how those fears could be classified into ten domains. It mentioned Average Fear Scores for them, ranging from a high of 2.15 for Man-Made Disasters to a low of 1.31 for Judgment of Others. But, it didn’t say where the individual Fear Scores landed in the range from 1 (Not Afraid) to 4 (Very Afraid). The Chapman University researchers had to calculate those individual scores before the averages, so why didn’t they mention the highest one? Where was that fear score for Corruption? Was it way up at 3.75, enough to make us Scream like that man in the famous painting by Edward Munch. Or was it 3.5, or 3.25, or just 3.0 (Afraid).

A fear score can be calculated from the Valid Percent answers for each question tabulated in the codebook of Complete Survey Results. The formula simply is a weighted average of the proportions:

Fear Score = [ 1x(% for Not Afraid) 
+  2x(% for Slightly Afraid) 
+ 3x(% for Afraid) 
+ 4x(% for Very Afraid)]/100  


I calculated all the the Fear Scores. Three tables show results for the top half of the list. (You can click on one to see a larger, clearer view).I have also shown their rank based on the Sum of [Very Afraid + Afraid] used by Chapman. For Corruption the score is only 2.7, which isn’t even at Afraid. For Public speaking it is 2.022, almost exactly 2.0 (or just Slightly Afraid). For Zombies it was 1.308, and the very lowest for Gender was 1.201. Often the rankings based on Fear Score and  the Sum of [Very Afraid + Afraid] are similar. The same ten appear in both Top Tens, but in slightly different order. The biggest difference is for Drunk driver, which is ranked 21st by Fear Score but 31st on the Chapman list. Three numbers in the rightmost column are shown in orange. For these items the 5 Afraid or Very Afraid shown in the Chapman blog post list didn't match their Complete Survey Results, and differs slightly.

As shown above in a line graph, those fear scores cover a range from a low of 1.2 to a high of 2.7. There’s no data either in the upper pink or lower green rectangles. The data cover just 1.5, or half the total of 3.0 you might expect to find. The highest score of 2.7 is not very frightening at all. It reminded me of some lyrics in Katy Perry’s very popular song This Is How We Do, which are:

“It’s No Big Deal.
It’s No Big Deal.
It’s No Big Deal.
This Is No Big Deal!”


Both the Fear Score and the Sum of [Very Afraid + Afraid] are reasonable for summarizing the results. It is easy to imagine data where the story they tell about how flat or peaked the data looks is incomplete, like the four symmetrical examples shown above which were constructed with a Sum of 50% and a Fear Score of 2.5. 

Another way to display how flat or peaked the data looks is by charting Very Afraid, Sum of [Very Afraid + Afraid], and Sum of [Very Afraid + Afraid + Slightly Afraid], as shown above. These three items also could be overlaid as a Stacked Bar Chart. 

Real data also can be skewed in either direction, like those for Corruption and Public Speaking shown above. You really have to look at the details. 

Fear Scores view the survey like a Fear Survey Schedule, a topic I last discussed on April 25th in a blog post titled Is public speaking by far the scariest thing that people face? Even more than death? No, it is not. A fear score describes what people fear most, rather than what most people fear. That distinction seems beyond the grasp of many journalists.   

No comments: